Title of Invention	A METHOD AND APPARATUS FOR TESTING A RADIO NETWORK UNDER INCREASED TRAFFIC LOAD
Abstract	In a radio communication system that includes multiple service areas (SAs), each service area (SA) is associated with a predetermined number or amount of radio communication resources. Each radio communication resource can be used for a prescribed period of time corresponding to one or more time units. An operator or other entity identifies (S1) a set of service areas to be tested at an increased effective load. A desired test level (e.g., increased traffic load) is determined (S2), and a corresponding number of radio resource time units is determined to achieve the desired test level conditions (S3 and S4). Use of radio communication resources in the determined one or more time units is blocked (S5) for a test period. Performance by each service area during the test period is determined (S6).

Title of Invention

A METHOD AND APPARATUS FOR TESTING A RADIO NETWORK UNDER INCREASED TRAFFIC LOAD

Abstract

In a radio communication system that includes multiple service areas (SAs), each service area (SA) is associated with a predetermined number or amount of radio communication resources. Each radio communication resource can be used for a prescribed period of time corresponding to one or more time units. An operator or other entity identifies (S1) a set of service areas to be tested at an increased effective load. A desired test level (e.g., increased traffic load) is determined (S2), and a corresponding number of radio resource time units is determined to achieve the desired test level conditions (S3 and S4). Use of radio communication resources in the determined one or more time units is blocked (S5) for a test period. Performance by each service area during the test period is determined (S6).

Full Text	FIELD OF THE INVENTION The invention relates to a system for testing cellular radio communication networks, and in particular, the performance of such networks under increased load. BACKGROUND AND SUMMARY The rapid, world-wide expansion of cellular networks and the introduction of new wireless services combined with competition among network operators has meant an ever-increasing need for continuous improvement as to quality, capacity, and accessibility. From the network operator perspective, higher quality, increased capacity, and better accessibility must be provided while also keeping the cost of calls and other services as low as possible. Moreover, to allow for future traffic growth and the introduction of new services, major investments in network equipment and functionality are typically necessary. This new equipment and functionality must be verified in realistic circumstances before being deployed commercially. Even with equipment and functionality already in place, it is important that operators can obtain a "proof of performance" with the possibility to identify and remedy problem areas before the higher loads and/or new services exist in commercial operation. Consequently, operators are particularly interested in obtaining information regarding the likely performance of a particular network under increased load conditions in which a greater percentage of the available radio resources is being used. For example, a network operator may want proof or demonstration regarding whether and how well an existing site configuration can provide additional service(s). But accurately providing this kind of information is difficult, particularly if there are insufficient users available to load those sites to the increased level desired and/or there is insufficient hardware currently installed at the sites to support the higher traffic load. To test the capacity of a particular network or site configuration that includes a plurality of sites and/or sectors, the traffic could be increased by simply increasing the number of people making calls in the test area, assuming that there is sufficient installed equipment to handle the higher load. But this kind of manual loading process is time- consuming and expensive and requires that a large number of people be employed and sent out to load up the network with calls. Another problem with the manual loading approach is that it is difficult to ensure that these newly-added test users mimic the behavior of real traffic loading since they are being asked to make "artificial" test calls. Their mobility patterns and cellular phone usage may significantly differ from those of real users, thereby raising doubts about the accuracy of the system performance results so obtained. An alternative approach to increase the effective load on the network test area would be to increase the radio resource burden of each existing user, for example, by disabling power control and/or discontinuous transmission (DTX). Since features such as power control and DTX reduce the power transmitted by each user, disabling them is equivalent to adding more users manually in terms of the traffic load level in the network. An advantage with this method over the manual loading technique is that the drawbacks regarding time, cost, organization, and accuracy of results outlined above are avoided. A disadvantage, however, is that the gain from features like power control and DTX is typically difficult to quantify in practical situations, and therefore, the effective load on the network achieved by disabling these features is uncertain. A better loading approach to increase the effective load on the network test area is to reduce the number of available radio resources. Traffic load is typically distributed onto a limited amount or number of radio resources. For example, in the context of a radio communications network that employs time division multiple access (TDMA) technology, the radio resources include time slots and frequencies. If the same number of users may only utilize a reduced quantity of radio resources, then the load on these radio resources is increased. If the radio resources in the system, e.g., time slots and frequencies in a TDMA system, are equivalent and independent, then the performance results obtained using the reduced, sub-set of radio resources can be extrapolated to give network performance measures for the full network resource situation at higher loads. A difficulty with reducing the amount or number of available radio resources in order to increase the effective load is that not all radio resources are equivalent and independent, which makes extrapolation of the test results uncertain. For example, frequency bandwidth is typically an important radio resource. But because multi-path fading is frequency-dependent, reducing the frequency bandwidth influences the ability of users to combat multi-path fading, which adversely affects performance. Hence, obtaining test results with a reduced frequency bandwidth suffers from the same uncertainties in extrapolation to overall network performance as those discussed above in the context of increasing the radio resource burden of each user, for example, by disabling power control and/or DTX. In the GSM TDMA system, users typically utilize frequency bandwidth by frequency hopping over multiple frequencies, each having 200kHz bandwidth. In a system that implements frequency hopping, like GSM, reducing the number of available radio resources might correspond to reducing the number of hopping frequencies. The result is that existing calls must be handled using the reduced number of frequencies, which increases the load on those remaining frequencies. A drawback with reducing the number of frequencies, particularly in a frequency hopping context, is that it adversely affects the ability to combat multi-path fading by reducing the frequency bandwidth used by a connection, as explained above. It also reduces the variation of radio quality within a radio block, which reduces decoding performance when significant channel coding is present, as is the case with GSM speech. Further, it reduces the interference averaging effect that allows the gains of some users, e.g. due to DTX, to benefit all users. Hence, increasing the effective load in a GSM network for test purposes by reducing the number of hopping frequencies has significant disadvantages because the radio environment experienced by the users is fundamentally altered in the process. In a frequency hopping GSM system, the effective load on the radio resources can be measured by the frequency load, which is defined as the served traffic (the number of users and their bandwidth requirements), divided by the number of hopping frequencies times the number of time slots. Since increasing the traffic via artificially adding more "test" users, via increasing the radio resource burden of the existing users, or via reducing the number of hopping frequencies all have disadvantages, a better way of increasing the load in a GSM system is to reduce the number of time slots in each frequency or frequency hopping channel group, for example, by blocking a predetermined number of time slots to traffic. Individual time slots are independent of the other time slots, and the radio environment experienced by the users is unaffected since the correct number of hopping frequencies is still used. By carefully selecting the number of time slots to be used, the frequency load can be increased without resulting in congestion to existing users. This is achieved by either ensuring that sufficient equipment is in place to prevent congestion in the frequency hopping channel group with the reduced number of time slots, or by creating an extra frequency or frequency hopping channel group with a full set of time slots that can serve users that would otherwise be denied access to the network. In GSM, such an extra frequency channel group may typically contain a non-frequency hopping, broadcast control channel (BCCH) frequency. This approach to increasing the effective load may be used in cellular networks that do not employ TDMA and/or frequency hopping. For example, the invention can be applied to orthogonal frequency division multiplexing (OFDM) and related access techniques by limiting the time of use for one or more sub-channel frequencies, as well as to spread spectrum, code division multiple access (CDMA) based systems. In all radio resource access techniques, time is a common radio resource. If desired, the approach can also be combined with other mechanisms/techniques for increasing effective system load, such as (but not limited to) those outlined above. The reduced time slot approach may be used in a radio communication system that includes multiple service areas. Each service area is associated with a predetermined number or amount of radio communication resources. Each radio communication resource can be used for a prescribed period of time which can be set by one or more time units. An operator or other entity identifies a set of service areas to be tested at an increased effective load. A desired test level (e.g., increased traffic load) is determined, and a corresponding number of radio resource time units is determined to achieve the desired test level conditions. Use of radio communication resources in the determined one or more time units is blocked for a test period. Performance in each service area during the test period is determined, and thereafter, aggregated into overall test network statistics as deemed appropriate. The performance may be determined based on measures such as: dropped call rate, received signal strength, signal quality, interference, handover success rate, and bit and block error rates. Each time unit may correspond to a time slot or a time frame, and each radio resource may be associated with a frequency, a frequency range, or a frequency hopping group. Alternatively, if the radio communication system uses code division multiple access (CDMA), each radio resource may be associated with the code, and each time unit may correspond to a transmission time interval (TTI). In that case, the radio transmitters in the set of service areas transmit at an increased power level during unblocked TTIs and at a decreased power level during blocked TTIs. In an example embodiment, synchronous time units in all of the service areas being tested are blocked. Synchronous time units or time slots are preferentially chosen because interference from one service area typically affects performance in other service areas. If the effect of the increased effective load is to be registered equally across the network test area in the form of increased interference, the load must be concentrated to the same time unit(s) in all service areas. In practice, the blocking of exactly synchronous time units or time slots may not be possible. This could be the case, for example, if different service areas obtain their timing reference from different and independent transmission links. In asynchronous systems, time units or time slots as close to synchronous as possible should be blocked in the different service areas. Any non-alignment can then be corrected for in the effective load calculation. Such a correction is desirable since non-alignment reduces the interference experienced by the traffic in the network, and therefore, the actual effective load achieved during the test. In one non-limiting, example embodiment applied to a control node, a controller in that node controls plural radio base station units. The radio base station units may correspond to base station sites or to base station sectors. Once a set of service areas is specified for testing, one or more times in each service area is determined when one or more radio resources associated with each such service area will not be used. A performance measure is determined under those conditions. Another non-limiting, example embodiment is a computer program product that includes computer code operable to control a computer. That code may include first logic operable to determine a test set of service areas for testing, second logic operable to determine one or more times in each service area when one or more radio resources for each service area will not be used, and third logic operable to determine each service area's performance under those conditions. BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 illustrates a radio communications system that includes a number of service areas, a set of which are to be tested; Figure 2 illustrates time slot resources associated with each of the sites to be tested in a synchronous system; Figure 3 illustrates time slots of the sites to be tested in an asynchronous system; Figure 4 is a flow chart diagram illustrating a non-limiting set of procedures for measuring the performance of certain sites under increased effective load; Figure 5 illustrates a controller for controlling the sites and controlling the test procedures; and Figures 6A and 6B illustrate drawings in which the transmit power in a particular site is concentrated to specific transmission time intervals. DETAILED DESCRIPTION The following description sets forth specific details, such as particular embodiments, procedures, techniques, etc., for purposes of explanation and not limitation. But it will be appreciated by one skilled in the art that other embodiments may be employed apart from these specific details. For example, although the following description makes reference to GSM-type radio resources, including time slots and frequency hopping, the invention may be employed in virtually any type of radio communication system in which time is a factor in allocating and using radio resources. In some instances, detailed descriptions of well-known methods, interfaces, circuits, etc., are omitted so as not to obscure the description with unnecessary detail. Moreover, individual blocks are shown in some of the figures. Those skilled in the art will appreciate that the functions of those blocks may be implemented using individual hardware circuits, using software program products in conjunction with a suitably-programmed microprocessor or other computer, using application specific circuitry (ASIC), and/or using one or more digital signal processors (DSPs). Figure 1 illustrates a radio communication system comprised of multiple service areas, referred to here generically as "sites", including sites S1 through S12. The term "site" encompasses a cell, a radio base station, or a radio base station sector. An effective load/traffic capacity test is desired to be performed on a certain set of service areas. In the example in Figure 1, sites S1 through S12 are preferably all put under test conditions to determine the performance of sites S1-S3. That is why sites S1-S3 are referred to as first-tier sites, and sites S4-S12 are referred to as second-tier sites. The second-tier sites are tested to include their effects, e.g., interference, etc., on the first-tier sites. For purposes of this example only, the radio communication system is assumed to be a TDMA type system that supports frequency hopping between a number of hopping frequencies, with each frequency being divided into repeating frames of eight time slots TS0-TS7. Figure 2 illustrates a time slot frame for each of sites S1-S3. In this case, the time slots are synchronized in time so that each time slot on each frequency starts and ends at the same time. In order to increase the frequency load to a predetermined value, communication using time slots TSO, TS1, and TS7 is blocked during a test period for each site S1-S1 2. The number of blocked time slots depends on the load level to be tested, and preferably, also takes into account the actual load currently being generated by subscribers. The load should be increased so that those subscribers do not detect congestion. Maintaining congestion at acceptable levels can be assured by installing sufficient equipment in the frequency hopping channel group with the reduced number of time slots, or by creating an extra frequency or frequency hopping channel group with enough time slots that can serve users that would otherwise be denied access to the network. As a result of blocking time slots TSO, TS1, and TS7, all of the traffic must be carried using only time slots TS2-TS6. This effectively increases the traffic load at each site S1-S12 and permits performance testing at sites S1-S3 under this increased load condition. Performance may be determined or measured in any suitable manner including (but not limited to) measuring a number of dropped calls, a dropped call rate, signal quality, signal strength, interference, bit or block error rates, delay, etc. Such performance indicators from each site are then aggregated into an overall network performance measure(s) as deemed appropriate by the operator. For example, a dropped call rate for sites S1-S3 and an average frequency load for sites S1-S12 may be calculated and compared to the operator- defined targets for traffic load and dropped call rate in the network. Furthermore, individual site data can be used to identify problem areas that require particular attention with respect to continued network optimization, site additions, equipment upgrades, and so on, in order for the operator to meet network capacity and performance targets. Conducting a similar performance test in an unsynchronized system is more complicated. Figure 3 shows such an unsynchronized system with the time slots for each of the sites S1-S3 being offset from each other. Time slots most closely aligned in each site are the best candidates for blocking. Exemplary time slots for blocking in this case are time slots TS5-TS7 for site S1, time slots TSO, TS1, and TS7 for site S2, and time slots TSO, TS6, and TS7 for site S3, as shown. In this way, the traffic load may be concentrated in five time slots in each of the sites that are as close to synchronous as possible, although they have different time slot numbers in the repeating frame structure of each site. The corresponding time slots in the second tier sites S4-S12 that are closely aligned to the blocked time slots in the first tier sites S1-S3 are also blocked. As with the synchronized case, the number of time slots blocked may be balanced with other factors such as avoiding congestion for active subscribers in the network and maintaining reasonable quality of service, in addition to the primary aim of reaching a target effective load for network testing. In order to ascertain the relative timing offsets of sites in an unsynchronized system, timing measurements must be performed, typically by using a test mobile station to register timing synchronization information broadcast by each site. This timing information for each site can be compared to determine offsets between the time slots at each of the sites, and thereby, which time slot numbers should be blocked in each site to achieve close to synchronous time slot blocking across the test area as desired. The frequency load of sites S1-S12 determined in an unsynchronized setting is preferably corrected for the lack of synchronization between time slots in the different sites. Such a correction is desirable since unaligned time slots reduce the interference experienced by traffic in the unblocked time slots, and therefore, the actual effective load achieved during a test period. One non-limiting, example way in which the measurements in an unsynchronized application may be corrected is now described. The maximum offset between corresponding blocked time slots in an arbitrary site and the chosen reference site is plus or minus one-half a time slot, if the time slots to block have been chosen to minimize their non-alignment. Hence, the maximum offset between any arbitrary site pair in the set is one time slot. A correction can be applied to the calculated frequency load based on a fixed, average timing offset in the site set. For example, it can be shown that the average timing offset between an arbitrary site pair with independent random offsets from the reference site in the range of-0.5 time slots to +0.5 time slots is 0.333 (one-third) time slots. Assume now as an example that two time slots in each site should be blocked in order to achieve the desired frequency load for the network test. The frequency load carried by sites with six unblocked time slots may then be reduced by a factor 0.333/6 when calculating the effective frequency load. This factor corresponds to the relative amount of traffic in the site that, on average, does not contribute to interference in the other sites. In some cases, in both synchronized and unsynchronized systems, it may be necessary to block fewer than the desired number of time slots in certain sites, for example, in order to prevent congestion if sufficient equipment to cater for the higher effective load has not been able to be installed. No correction is necessary for sites with fewer than the maximum number of blocked time slots. This is because the additional unblocked time slot(s) can always be chosen so as to cover the offset to other sites that is at most one full time slot, as explained above. As a result, there is no interference reduction to the core set of unblocked time slots, i.e., those common to all sites, whose performance is to be evaluated. By way of illustration, consider the following, non-limiting, frequency-load formulas for a scenario where two time slots have been blocked. In sites with eight unblocked time slots: frequency load (FL) = Erlang/ (6 * 8). In sites with seven unblocked time slots: FL = Erlang/ (6 * 7) In sites with six unblocked time slots: FL = (1 -0.333/6)* Erlang/ (6 * 6) The total frequency load in the set of test sites is as follows: FL (total) = (Σ FL in all sites in the set)/(number of sites in the set). Figure 4 illustrates an example, non-limiting set of procedures for implementing site testing under increased effective load. A set of multiple service areas (MSAs) is determined for testing (step S1). In the example above, that testing is conducted in the context of increased frequency load. Again, a service area may be a site, a sector, a cell, or any other type of area. Based on the desired test load and the current load in the service areas, the number of time slots in each service area to be blocked during the test period is determined, preferably also ensuring that congestion can be avoided (step S2). If the service areas are not synchronized, the timing signals broadcast from each service area are measured, for example, by using a roaming mobile "test" receiver. That timing and time slot number information is conveyed to a control node for later use in identifying the time slots to be blocked in the service areas and in correcting the effective load for any time slot non-alignment. If the service areas are synchronized, the control node typically already knows the timing and time slot number information. Otherwise, it can obtain the information directly from the radio base station associated with that service area (step S3). Based on the timing and time slot number information and the number of time slots to be blocked from step S2, a determination can be made as to which time slots are synchronous and should be blocked in all of the mobile service areas, or in an unsynchronized context, which time slots are the most closely-aligned and should be blocked, using the timing information from an arbitrary service area as a reference (step S4). Those time slots identified in step S4 are blocked (step S5). Performance is then measured or otherwise determined under the increased load condition with certain time slots blocked. For unsynchronized systems, the effective load is corrected to compensate for unaligned time slots (step S6). Figure 5 illustrates a controller node 10 that includes a CPU 12 and a memory 14 with a program that stores logic code to implement the steps illustrated, for example, in Figure 4. Of course, other steps and procedures may be coded to implement the claimed invention. The controller 10 in Figure 5 is coupled to each of the sites S1-S12, consistent with the sites shown in the two tier structure of Figure 1, where sites S1-S3 are first-tier sites and sites S4-S12 are second-tier sites. Although each radio resource may also be associated with a frequency, frequency range, or frequency hopping group, such as illustrated in Figures 2 and 3 in the context of a TDMA-type system, the radio resource may also correspond, for example, to a spreading code like those used in a code division multiple access system. In a CDMA system, each time unit then corresponds to a transmission time interval (TTI). Reference is made to Figures 6A and 6B which illustrate the transmission power in a particular site in Figure 6. Figure 6A is a graph that shows a normal transmission situation in which the power level (normalized to "1") is consistent over consecutive TTIs. However, to test the capacity of the system, signals may be transmitted in a "compressed mode," an example of which is illustrated in Figure 6B. The transmitters in each site being tested transmit at twice the power during a first transmission time interval, but at zero power in the following second TTI. This pattern effectively doubles the load in the service area during alternating TTIs, and the performance of the system can be measured during those increased transmit power TTIs. Of course, the power may be increased by something other than twice (either more or less), during the first TTI, and the power during the second TTI may be something more than zero. Also, both the higher and lower power periods may extend for more than one TTI, and they need not extend for an equal number of TTIs. As these figures illustrate, an important radio resource in a CDMA system is transmission power, which is allocated with spreading codes. Here, the time unit being blocked corresponds to alternating TTIs which results in increased traffic load during the other TTIs. There may be situations in which it is desirable to use the time unit/slot blocking approach in addition to one or more other load increasing strategies. Such other strategies might include (but are not limited to) decreasing a number of frequencies for use, increasing a number of mobile users, turning off discontinuous transmission (DTX), turning off power control, etc. A combination of loading methods may be desirable when the target effective load for testing is far greater than the current load in the test area. Many time slots must then be blocked to achieve the requisite load, which makes congestion harder to avoid in practice since significant additional equipment beyond that required for the current network load must be in place to do so. A combination of loading techniques may then become a more feasible alternative. The time unit blocking approach to increasing the effective load may be used with other cellular network technologies than those based on TDMA and CDMA, because time is always an important resource. For example, in orthogonal frequency division multiplexing (OFDM) and related access techniques, the time of use for one or more sub- channel frequencies, so-called tones, can be limited in order to increase the effective system load. With the primary radio resources being frequency and time, OFDM systems bear strong similarities in this context to the TDMA based systems like GSM discussed in detail above. The time unit blocking approach achieves numerous advantages over other techniques. First, the operator can test the network under increased effective load conditions without having to manually add extra users to the network, which saves time and money and gives truly representative performance results. Second, the operator can maintain the true radio resource burden of the users without having to disable features like power control and DTX in order to achieve an increased effective load whose magnitude is then difficult to quantify. Third, the operator can increase the effective system load without changing the radio environment for the users, which would be the case if the frequency bandwidth and/or the number of hopping frequencies were reduced, thereby ensuring representative performance results. Fourth, congestion can be avoided by ensuring that sufficient equipment is installed, by creating an extra channel group containing a full set of time units to support users that would otherwise be denied access to the network, by blocking fewer than the maximum number of time units in certain sites where congestion is otherwise deemed likely, or by some combination of the above. Fifth, the time unit blocking approach can be combined with other techniques for increasing the effective system load if desired. Sixth, the time unit blocking approach works in principle for all cellular network technologies because time is always a key resource. Seventh, the time unit blocking approach is applicable in both synchronous and asynchronous cellular systems. Eighth, the time unit blocking approach is easily implemented and operated in a control node in which the test preparations, the test itself, and the subsequent results analysis, both on a network and local site level, can be performed with minimal impact to normal network operations. While practical and preferred implementations of example embodiments have been described, it is to be understood that the invention is not limited to any disclosed embodiment or implementation, and on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims. WE CLAIM 1. A method for testing a radio network under increased load, which method is use in a radio communications system including multiple service areas, SAs, (SI-S12), each SA including a predetermined number of radio communication resources, where each radio communication resource is associated with a time unit, comprising identifying (S 1) a set of SAs for testing a desired test parameter or level, the method characterized by: Determining (S2) one or more radio communication resource, RCR, time units in each SA in the set of SAs corresponding to the desired test parameter or level; blocking (S5)use of radio communication resources in the determined one or more RCR time units for a test period; and determining (S6) a performance associated with the set of SAs during the test period. 2. The method in claim 1, wherein each radio communication resource is associated with a frequency, a frequency range, or a frequency hopping group. 3. The method in claim 1, wherein each time unit corresponds to a time slot or time frame. 4. The method in claim 1, wherein the test parameter level is traffic load or traffic capacity and the performance includes one or more of: a number of dropped calls, a dropped call rate, signal strength, signal quality, interference, delay, and an error rate. 5. The method in claim 1, wherein the radio communications system uses code division multiple access, CDMA, each radio communication resource is associated with a code, and each time unit corresponds to a transmission time interval, TTI. 6. The method in claim 5, wherein radio transmitters in the set of SAs transmit at an increased power during unblocked TTIs and at decreased power during blocked TTIs. 7. The method in claim 1, wherein the time units in each of the SAs in the set are synchronized. 8. The method in claim 1, wherein the time units in each of the SAs in the set are unsynchronized, the method comprising: determining which of the RCR time units for each of the SAs are most closely aligned, wherein the one or more blocked time units includes one or more of the determined RCR time units. 9. The method in claim 8 comprising: compensating the desired test parameter or the determined performance for a lack of synchronization between RCR time units in each of the SAs. 10. The method in claim 1 comprising : increasing a traffic load on the set of SAs during the test period in addition to blocking one or more RCR time units. 11. The method in claim 1 comprising: determining a number of RCR time units to block that avoids or limits congestion for active subscribers below a predetermined level. 12. A control node (10) for use in a radio communications system for controlling radio base station, RBS, units, each RBS unit associated with a service area (S1-S12) and each service area including a predetermined number of radio resources, comprising electronic circuitry (12, 14) configured to specify a set of service areas (S1-S12) for testing, characterized by the electronic circuitry (12, 14) being further configured to: determine one or more times in each service area (SI-SI2) when one or more radio resources associated with each service area (S1-S 12) will not be used; and determine performance when the one or more times in each service area when one or more radio resources for each service area is not used. 13. The control node in claim 12, wherein the control node (10) is an RBS controller. 14. The control node in claim 12, wherein the RBS units correspond to base station sites or to base station sectors. 15. The control node in claim 12, wherein each radio resource is associated with a frequency, a frequency range, or a frequency hopping group. 16. The control node in claim 12, wherein each of the one or more times corresponds to a time slot or time frame. 17. The control node in claim 12, wherein the performance is determined based on traffic load or traffic capacity using one or more of: a number of dropped calls, dropped call rate, signal strength, signal quality, interference, delay, and an error rate. 18. The control node in claim 12, wherein the radio communications system uses code division multiple access ,CDMA, each radio resource is associated with a code, and each time corresponds to a transmission time interval, TTI,. 19. The control node in claim 18, wherein the electronic circuitry is configured to instruct radio transmitters in the set of service areas to transmit at an increased power during unblocked TTIs and at decreased power during blocked TTIs. 20. The control node in claim 12, wherein the times in each of the service areas in the set are synchronized. 21. The control node in claim 12, wherein the time units in each of the service areas in the set are unsynchronized, the electronic circuitry is further configured to: determine which of the radio resource times for each of the service areas in the set are most closely aligned, wherein the one or more blocked times includes one or more of the determined times. 22. The control node in claim 12, wherein the electronic circuitry is further configured to: compensate the desired test parameter or the determined performance for a lack of synchronization between times in each of the service areas. 23. The control node in claim 12, wherein the electronic circuitry is further configured to: increasing a load on the set of service areas during performance determination in addition to blocking one or more radio resource times. 24. The control node in claim 12, wherein the electronic circuitry is further configured to: determine a number of times not to use that avoids or limits congestion for active subscribers below a predetermined level. ABSTRACT A METHOD AND APPARATUS FOR TESTING A RADIO NETWORK UNDER INCREASED TRAFFIC LOAD In a radio communication system that includes multiple service areas (SAs), each service area (SA) is associated with a predetermined number or amount of radio communication resources. Each radio communication resource can be used for a prescribed period of time corresponding to one or more time units. An operator or other entity identifies (S1) a set of service areas to be tested at an increased effective load. A desired test level (e.g., increased traffic load) is determined (S2), and a corresponding number of radio resource time units is determined to achieve the desired test level conditions (S3 and S4). Use of radio communication resources in the determined one or more time units is blocked (S5) for a test period. Performance by each service area during the test period is determined (S6).

Full Text

FIELD OF THE INVENTION
The invention relates to a system for testing cellular radio communication
networks, and in particular, the performance of such networks under increased load.
BACKGROUND AND SUMMARY
The rapid, world-wide expansion of cellular networks and the introduction of
new wireless services combined with competition among network operators has meant an
ever-increasing need for continuous improvement as to quality, capacity, and accessibility.
From the network operator perspective, higher quality, increased capacity, and better
accessibility must be provided while also keeping the cost of calls and other services as low
as possible. Moreover, to allow for future traffic growth and the introduction of new
services, major investments in network equipment and functionality are typically necessary.
This new equipment and functionality must be verified in realistic circumstances before being
deployed commercially. Even with equipment and functionality already in place, it is
important that operators can obtain a "proof of performance" with the possibility to identify
and remedy problem areas before the higher loads and/or new services exist in commercial
operation.
Consequently, operators are particularly interested in obtaining information
regarding the likely performance of a particular network under increased load conditions in
which a greater percentage of the available radio resources is being used. For example, a
network operator may want proof or demonstration regarding whether and how well an
existing site configuration can provide additional service(s). But accurately providing this
kind of information is difficult, particularly if there are insufficient users available to load
those sites to the increased level desired and/or there is insufficient hardware currently
installed at the sites to support the higher traffic load.
To test the capacity of a particular network or site configuration that includes a
plurality of sites and/or sectors, the traffic could be increased by simply increasing the
number of people making calls in the test area, assuming that there is sufficient installed
equipment to handle the higher load. But this kind of manual loading process is time-
consuming and expensive and requires that a large number of people be employed and sent
out to load up the network with calls. Another problem with the manual loading approach is

that it is difficult to ensure that these newly-added test users mimic the behavior of real traffic
loading since they are being asked to make "artificial" test calls. Their mobility patterns and
cellular phone usage may significantly differ from those of real users, thereby raising doubts
about the accuracy of the system performance results so obtained.
An alternative approach to increase the effective load on the network test area
would be to increase the radio resource burden of each existing user, for example, by
disabling power control and/or discontinuous transmission (DTX). Since features such as
power control and DTX reduce the power transmitted by each user, disabling them is
equivalent to adding more users manually in terms of the traffic load level in the network.
An advantage with this method over the manual loading technique is that the drawbacks
regarding time, cost, organization, and accuracy of results outlined above are avoided. A
disadvantage, however, is that the gain from features like power control and DTX is typically
difficult to quantify in practical situations, and therefore, the effective load on the network
achieved by disabling these features is uncertain.
A better loading approach to increase the effective load on the network test
area is to reduce the number of available radio resources. Traffic load is typically distributed
onto a limited amount or number of radio resources. For example, in the context of a radio
communications network that employs time division multiple access (TDMA) technology,
the radio resources include time slots and frequencies. If the same number of users may only
utilize a reduced quantity of radio resources, then the load on these radio resources is
increased. If the radio resources in the system, e.g., time slots and frequencies in a TDMA
system, are equivalent and independent, then the performance results obtained using the
reduced, sub-set of radio resources can be extrapolated to give network performance
measures for the full network resource situation at higher loads.
A difficulty with reducing the amount or number of available radio resources
in order to increase the effective load is that not all radio resources are equivalent and
independent, which makes extrapolation of the test results uncertain. For example, frequency
bandwidth is typically an important radio resource. But because multi-path fading is
frequency-dependent, reducing the frequency bandwidth influences the ability of users to
combat multi-path fading, which adversely affects performance. Hence, obtaining test results
with a reduced frequency bandwidth suffers from the same uncertainties in extrapolation to
overall network performance as those discussed above in the context of increasing the radio
resource burden of each user, for example, by disabling power control and/or DTX.

In the GSM TDMA system, users typically utilize frequency bandwidth by
frequency hopping over multiple frequencies, each having 200kHz bandwidth. In a system
that implements frequency hopping, like GSM, reducing the number of available radio
resources might correspond to reducing the number of hopping frequencies. The result is that
existing calls must be handled using the reduced number of frequencies, which increases the
load on those remaining frequencies. A drawback with reducing the number of frequencies,
particularly in a frequency hopping context, is that it adversely affects the ability to combat
multi-path fading by reducing the frequency bandwidth used by a connection, as explained
above. It also reduces the variation of radio quality within a radio block, which reduces
decoding performance when significant channel coding is present, as is the case with GSM
speech. Further, it reduces the interference averaging effect that allows the gains of some
users, e.g. due to DTX, to benefit all users. Hence, increasing the effective load in a GSM
network for test purposes by reducing the number of hopping frequencies has significant
disadvantages because the radio environment experienced by the users is fundamentally
altered in the process.
In a frequency hopping GSM system, the effective load on the radio resources
can be measured by the frequency load, which is defined as the served traffic (the number of
users and their bandwidth requirements), divided by the number of hopping frequencies times
the number of time slots. Since increasing the traffic via artificially adding more "test" users,
via increasing the radio resource burden of the existing users, or via reducing the number of
hopping frequencies all have disadvantages, a better way of increasing the load in a GSM
system is to reduce the number of time slots in each frequency or frequency hopping channel
group, for example, by blocking a predetermined number of time slots to traffic. Individual
time slots are independent of the other time slots, and the radio environment experienced by
the users is unaffected since the correct number of hopping frequencies is still used. By
carefully selecting the number of time slots to be used, the frequency load can be increased
without resulting in congestion to existing users. This is achieved by either ensuring that
sufficient equipment is in place to prevent congestion in the frequency hopping channel
group with the reduced number of time slots, or by creating an extra frequency or frequency
hopping channel group with a full set of time slots that can serve users that would otherwise
be denied access to the network. In GSM, such an extra frequency channel group may
typically contain a non-frequency hopping, broadcast control channel (BCCH) frequency.

This approach to increasing the effective load may be used in cellular
networks that do not employ TDMA and/or frequency hopping. For example, the invention
can be applied to orthogonal frequency division multiplexing (OFDM) and related access
techniques by limiting the time of use for one or more sub-channel frequencies, as well as to
spread spectrum, code division multiple access (CDMA) based systems. In all radio resource
access techniques, time is a common radio resource. If desired, the approach can also be
combined with other mechanisms/techniques for increasing effective system load, such as
(but not limited to) those outlined above.
The reduced time slot approach may be used in a radio communication system
that includes multiple service areas. Each service area is associated with a predetermined
number or amount of radio communication resources. Each radio communication resource
can be used for a prescribed period of time which can be set by one or more time units. An
operator or other entity identifies a set of service areas to be tested at an increased effective
load. A desired test level (e.g., increased traffic load) is determined, and a corresponding
number of radio resource time units is determined to achieve the desired test level conditions.
Use of radio communication resources in the determined one or more time units is blocked
for a test period. Performance in each service area during the test period is determined, and
thereafter, aggregated into overall test network statistics as deemed appropriate.
The performance may be determined based on measures such as: dropped call
rate, received signal strength, signal quality, interference, handover success rate, and bit and
block error rates. Each time unit may correspond to a time slot or a time frame, and each
radio resource may be associated with a frequency, a frequency range, or a frequency
hopping group. Alternatively, if the radio communication system uses code division multiple
access (CDMA), each radio resource may be associated with the code, and each time unit
may correspond to a transmission time interval (TTI). In that case, the radio transmitters in
the set of service areas transmit at an increased power level during unblocked TTIs and at a
decreased power level during blocked TTIs.
In an example embodiment, synchronous time units in all of the service areas
being tested are blocked. Synchronous time units or time slots are preferentially chosen
because interference from one service area typically affects performance in other service
areas. If the effect of the increased effective load is to be registered equally across the
network test area in the form of increased interference, the load must be concentrated to the
same time unit(s) in all service areas.

In practice, the blocking of exactly synchronous time units or time slots may
not be possible. This could be the case, for example, if different service areas obtain their
timing reference from different and independent transmission links. In asynchronous
systems, time units or time slots as close to synchronous as possible should be blocked in the
different service areas. Any non-alignment can then be corrected for in the effective load
calculation. Such a correction is desirable since non-alignment reduces the interference
experienced by the traffic in the network, and therefore, the actual effective load achieved
during the test.
In one non-limiting, example embodiment applied to a control node, a
controller in that node controls plural radio base station units. The radio base station units
may correspond to base station sites or to base station sectors. Once a set of service areas is
specified for testing, one or more times in each service area is determined when one or more
radio resources associated with each such service area will not be used. A performance
measure is determined under those conditions. Another non-limiting, example embodiment
is a computer program product that includes computer code operable to control a computer.
That code may include first logic operable to determine a test set of service areas for testing,
second logic operable to determine one or more times in each service area when one or more
radio resources for each service area will not be used, and third logic operable to determine
each service area's performance under those conditions.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 illustrates a radio communications system that includes a number
of service areas, a set of which are to be tested;
Figure 2 illustrates time slot resources associated with each of the sites to be
tested in a synchronous system;
Figure 3 illustrates time slots of the sites to be tested in an asynchronous
system;
Figure 4 is a flow chart diagram illustrating a non-limiting set of procedures
for measuring the performance of certain sites under increased effective load;
Figure 5 illustrates a controller for controlling the sites and controlling the test
procedures; and
Figures 6A and 6B illustrate drawings in which the transmit power in a
particular site is concentrated to specific transmission time intervals.

DETAILED DESCRIPTION
The following description sets forth specific details, such as particular
embodiments, procedures, techniques, etc., for purposes of explanation and not limitation.
But it will be appreciated by one skilled in the art that other embodiments may be employed
apart from these specific details. For example, although the following description makes
reference to GSM-type radio resources, including time slots and frequency hopping, the
invention may be employed in virtually any type of radio communication system in which
time is a factor in allocating and using radio resources. In some instances, detailed
descriptions of well-known methods, interfaces, circuits, etc., are omitted so as not to obscure
the description with unnecessary detail. Moreover, individual blocks are shown in some of
the figures. Those skilled in the art will appreciate that the functions of those blocks may be
implemented using individual hardware circuits, using software program products in
conjunction with a suitably-programmed microprocessor or other computer, using application
specific circuitry (ASIC), and/or using one or more digital signal processors (DSPs).
Figure 1 illustrates a radio communication system comprised of multiple
service areas, referred to here generically as "sites", including sites S1 through S12. The term
"site" encompasses a cell, a radio base station, or a radio base station sector. An effective
load/traffic capacity test is desired to be performed on a certain set of service areas. In the
example in Figure 1, sites S1 through S12 are preferably all put under test conditions to
determine the performance of sites S1-S3. That is why sites S1-S3 are referred to as first-tier
sites, and sites S4-S12 are referred to as second-tier sites. The second-tier sites are tested to
include their effects, e.g., interference, etc., on the first-tier sites. For purposes of this
example only, the radio communication system is assumed to be a TDMA type system that
supports frequency hopping between a number of hopping frequencies, with each frequency
being divided into repeating frames of eight time slots TS0-TS7.
Figure 2 illustrates a time slot frame for each of sites S1-S3. In this case, the
time slots are synchronized in time so that each time slot on each frequency starts and ends at
the same time. In order to increase the frequency load to a predetermined value,
communication using time slots TSO, TS1, and TS7 is blocked during a test period for each
site S1-S1 2. The number of blocked time slots depends on the load level to be tested, and
preferably, also takes into account the actual load currently being generated by subscribers.
The load should be increased so that those subscribers do not detect congestion. Maintaining
congestion at acceptable levels can be assured by installing sufficient equipment in the

frequency hopping channel group with the reduced number of time slots, or by creating an
extra frequency or frequency hopping channel group with enough time slots that can serve
users that would otherwise be denied access to the network. As a result of blocking time slots
TSO, TS1, and TS7, all of the traffic must be carried using only time slots TS2-TS6. This
effectively increases the traffic load at each site S1-S12 and permits performance testing at
sites S1-S3 under this increased load condition.
Performance may be determined or measured in any suitable manner including
(but not limited to) measuring a number of dropped calls, a dropped call rate, signal quality,
signal strength, interference, bit or block error rates, delay, etc. Such performance indicators
from each site are then aggregated into an overall network performance measure(s) as
deemed appropriate by the operator. For example, a dropped call rate for sites S1-S3 and an
average frequency load for sites S1-S12 may be calculated and compared to the operator-
defined targets for traffic load and dropped call rate in the network. Furthermore, individual
site data can be used to identify problem areas that require particular attention with respect to
continued network optimization, site additions, equipment upgrades, and so on, in order for
the operator to meet network capacity and performance targets.
Conducting a similar performance test in an unsynchronized system is more
complicated. Figure 3 shows such an unsynchronized system with the time slots for each of
the sites S1-S3 being offset from each other. Time slots most closely aligned in each site are
the best candidates for blocking. Exemplary time slots for blocking in this case are time slots
TS5-TS7 for site S1, time slots TSO, TS1, and TS7 for site S2, and time slots TSO, TS6, and
TS7 for site S3, as shown. In this way, the traffic load may be concentrated in five time slots
in each of the sites that are as close to synchronous as possible, although they have different
time slot numbers in the repeating frame structure of each site. The corresponding time slots
in the second tier sites S4-S12 that are closely aligned to the blocked time slots in the first tier
sites S1-S3 are also blocked. As with the synchronized case, the number of time slots
blocked may be balanced with other factors such as avoiding congestion for active
subscribers in the network and maintaining reasonable quality of service, in addition to the
primary aim of reaching a target effective load for network testing.
In order to ascertain the relative timing offsets of sites in an unsynchronized
system, timing measurements must be performed, typically by using a test mobile station to
register timing synchronization information broadcast by each site. This timing information
for each site can be compared to determine offsets between the time slots at each of the sites,

and thereby, which time slot numbers should be blocked in each site to achieve close to
synchronous time slot blocking across the test area as desired.
The frequency load of sites S1-S12 determined in an unsynchronized setting is
preferably corrected for the lack of synchronization between time slots in the different sites.
Such a correction is desirable since unaligned time slots reduce the interference experienced
by traffic in the unblocked time slots, and therefore, the actual effective load achieved during
a test period.
One non-limiting, example way in which the measurements in an
unsynchronized application may be corrected is now described. The maximum offset
between corresponding blocked time slots in an arbitrary site and the chosen reference site is
plus or minus one-half a time slot, if the time slots to block have been chosen to minimize
their non-alignment. Hence, the maximum offset between any arbitrary site pair in the set is
one time slot. A correction can be applied to the calculated frequency load based on a fixed,
average timing offset in the site set. For example, it can be shown that the average timing
offset between an arbitrary site pair with independent random offsets from the reference site
in the range of-0.5 time slots to +0.5 time slots is 0.333 (one-third) time slots. Assume now
as an example that two time slots in each site should be blocked in order to achieve the
desired frequency load for the network test. The frequency load carried by sites with six
unblocked time slots may then be reduced by a factor 0.333/6 when calculating the effective
frequency load. This factor corresponds to the relative amount of traffic in the site that, on
average, does not contribute to interference in the other sites.
In some cases, in both synchronized and unsynchronized systems, it may be
necessary to block fewer than the desired number of time slots in certain sites, for example, in
order to prevent congestion if sufficient equipment to cater for the higher effective load has
not been able to be installed. No correction is necessary for sites with fewer than the
maximum number of blocked time slots. This is because the additional unblocked time
slot(s) can always be chosen so as to cover the offset to other sites that is at most one full
time slot, as explained above. As a result, there is no interference reduction to the core set of
unblocked time slots, i.e., those common to all sites, whose performance is to be evaluated.
By way of illustration, consider the following, non-limiting, frequency-load
formulas for a scenario where two time slots have been blocked. In sites with eight
unblocked time slots:
frequency load (FL) = Erlang/ (6 * 8).

In sites with seven unblocked time slots:
FL = Erlang/ (6 * 7)
In sites with six unblocked time slots:
FL = (1 -0.333/6)* Erlang/ (6 * 6)
The total frequency load in the set of test sites is as follows:
FL (total) = (Σ FL in all sites in the set)/(number of sites in the set).
Figure 4 illustrates an example, non-limiting set of procedures for
implementing site testing under increased effective load. A set of multiple service areas
(MSAs) is determined for testing (step S1). In the example above, that testing is conducted in
the context of increased frequency load. Again, a service area may be a site, a sector, a cell,
or any other type of area. Based on the desired test load and the current load in the service
areas, the number of time slots in each service area to be blocked during the test period is
determined, preferably also ensuring that congestion can be avoided (step S2).
If the service areas are not synchronized, the timing signals broadcast from
each service area are measured, for example, by using a roaming mobile "test" receiver. That
timing and time slot number information is conveyed to a control node for later use in
identifying the time slots to be blocked in the service areas and in correcting the effective
load for any time slot non-alignment. If the service areas are synchronized, the control node
typically already knows the timing and time slot number information. Otherwise, it can
obtain the information directly from the radio base station associated with that service area
(step S3).
Based on the timing and time slot number information and the number of time
slots to be blocked from step S2, a determination can be made as to which time slots are
synchronous and should be blocked in all of the mobile service areas, or in an
unsynchronized context, which time slots are the most closely-aligned and should be blocked,
using the timing information from an arbitrary service area as a reference (step S4). Those
time slots identified in step S4 are blocked (step S5). Performance is then measured or
otherwise determined under the increased load condition with certain time slots blocked. For
unsynchronized systems, the effective load is corrected to compensate for unaligned time
slots (step S6).
Figure 5 illustrates a controller node 10 that includes a CPU 12 and a memory
14 with a program that stores logic code to implement the steps illustrated, for example, in
Figure 4. Of course, other steps and procedures may be coded to implement the claimed

invention. The controller 10 in Figure 5 is coupled to each of the sites S1-S12, consistent
with the sites shown in the two tier structure of Figure 1, where sites S1-S3 are first-tier sites
and sites S4-S12 are second-tier sites. Although each radio resource may also be associated
with a frequency, frequency range, or frequency hopping group, such as illustrated in Figures
2 and 3 in the context of a TDMA-type system, the radio resource may also correspond, for
example, to a spreading code like those used in a code division multiple access system. In a
CDMA system, each time unit then corresponds to a transmission time interval (TTI).
Reference is made to Figures 6A and 6B which illustrate the transmission
power in a particular site in Figure 6. Figure 6A is a graph that shows a normal transmission
situation in which the power level (normalized to "1") is consistent over consecutive TTIs.
However, to test the capacity of the system, signals may be transmitted in a "compressed
mode," an example of which is illustrated in Figure 6B. The transmitters in each site being
tested transmit at twice the power during a first transmission time interval, but at zero power
in the following second TTI. This pattern effectively doubles the load in the service area
during alternating TTIs, and the performance of the system can be measured during those
increased transmit power TTIs. Of course, the power may be increased by something other
than twice (either more or less), during the first TTI, and the power during the second TTI
may be something more than zero. Also, both the higher and lower power periods may
extend for more than one TTI, and they need not extend for an equal number of TTIs. As
these figures illustrate, an important radio resource in a CDMA system is transmission power,
which is allocated with spreading codes. Here, the time unit being blocked corresponds to
alternating TTIs which results in increased traffic load during the other TTIs.
There may be situations in which it is desirable to use the time unit/slot
blocking approach in addition to one or more other load increasing strategies. Such other
strategies might include (but are not limited to) decreasing a number of frequencies for use,
increasing a number of mobile users, turning off discontinuous transmission (DTX), turning
off power control, etc. A combination of loading methods may be desirable when the target
effective load for testing is far greater than the current load in the test area. Many time slots
must then be blocked to achieve the requisite load, which makes congestion harder to avoid
in practice since significant additional equipment beyond that required for the current
network load must be in place to do so. A combination of loading techniques may then
become a more feasible alternative.

The time unit blocking approach to increasing the effective load may be used
with other cellular network technologies than those based on TDMA and CDMA, because
time is always an important resource. For example, in orthogonal frequency division
multiplexing (OFDM) and related access techniques, the time of use for one or more sub-
channel frequencies, so-called tones, can be limited in order to increase the effective system
load. With the primary radio resources being frequency and time, OFDM systems bear
strong similarities in this context to the TDMA based systems like GSM discussed in detail
above.
The time unit blocking approach achieves numerous advantages over other
techniques. First, the operator can test the network under increased effective load conditions
without having to manually add extra users to the network, which saves time and money and
gives truly representative performance results. Second, the operator can maintain the true
radio resource burden of the users without having to disable features like power control and
DTX in order to achieve an increased effective load whose magnitude is then difficult to
quantify. Third, the operator can increase the effective system load without changing the
radio environment for the users, which would be the case if the frequency bandwidth and/or
the number of hopping frequencies were reduced, thereby ensuring representative
performance results. Fourth, congestion can be avoided by ensuring that sufficient equipment
is installed, by creating an extra channel group containing a full set of time units to support
users that would otherwise be denied access to the network, by blocking fewer than the
maximum number of time units in certain sites where congestion is otherwise deemed likely,
or by some combination of the above. Fifth, the time unit blocking approach can be
combined with other techniques for increasing the effective system load if desired. Sixth, the
time unit blocking approach works in principle for all cellular network technologies because
time is always a key resource. Seventh, the time unit blocking approach is applicable in both
synchronous and asynchronous cellular systems. Eighth, the time unit blocking approach is
easily implemented and operated in a control node in which the test preparations, the test
itself, and the subsequent results analysis, both on a network and local site level, can be
performed with minimal impact to normal network operations.
While practical and preferred implementations of example embodiments have
been described, it is to be understood that the invention is not limited to any disclosed
embodiment or implementation, and on the contrary, is intended to cover various
modifications and equivalent arrangements included within the scope of the appended claims.

WE CLAIM
1. A method for testing a radio network under increased load, which
method is use in a radio communications system including multiple service areas, SAs,
(SI-S12), each SA including a predetermined number of radio communication resources,
where each radio communication resource is associated with a time unit, comprising
identifying (S 1) a set of SAs for testing a desired test parameter or level, the method
characterized by:
Determining (S2) one or more radio communication resource, RCR, time units in
each SA in the set of SAs corresponding to the desired test parameter or level;
blocking (S5)use of radio communication resources in the determined one
or more RCR time units for a test period; and
determining (S6) a performance associated with the set of SAs during the test
period.
2. The method in claim 1, wherein each radio communication resource is associated
with a frequency, a frequency range, or a frequency hopping group.
3. The method in claim 1, wherein each time unit corresponds to a time slot or time
frame.
4. The method in claim 1, wherein the test parameter level is traffic load or traffic
capacity and the performance includes one or more of: a number of dropped calls, a
dropped call rate, signal strength, signal quality, interference, delay, and an error rate.
5. The method in claim 1, wherein the radio communications system uses code
division multiple access, CDMA, each radio communication resource is associated with a
code, and each time unit corresponds to a transmission time interval, TTI.
6. The method in claim 5, wherein radio transmitters in the set of SAs transmit at an
increased power during unblocked TTIs and at decreased power during blocked TTIs.
7. The method in claim 1, wherein the time units in each of the SAs in the set are
synchronized.

8. The method in claim 1, wherein the time units in each of the SAs in the set are
unsynchronized, the method comprising:
determining which of the RCR time units for each of the SAs are most closely
aligned,
wherein the one or more blocked time units includes one or more of the
determined RCR time units.
9. The method in claim 8 comprising:
compensating the desired test parameter or the determined performance for a
lack of synchronization between RCR time units in each of the SAs.
10. The method in claim 1 comprising :
increasing a traffic load on the set of SAs during the test period in addition to
blocking one or more RCR time units.
11. The method in claim 1 comprising:
determining a number of RCR time units to block that avoids or limits congestion
for active subscribers below a predetermined level.
12. A control node (10) for use in a radio communications system for controlling radio
base station, RBS, units, each RBS unit associated with a service area (S1-S12) and each
service area including a predetermined number of radio resources, comprising electronic
circuitry (12, 14) configured to specify a set of service areas (S1-S12) for testing,
characterized by the electronic circuitry (12, 14) being further configured to:
determine one or more times in each service area (SI-SI2) when one or more
radio resources associated with each service area (S1-S 12) will not be used; and
determine performance when the one or more times in each service area when
one or more radio resources for each service area is not used.
13. The control node in claim 12, wherein the control node (10) is an RBS controller.
14. The control node in claim 12, wherein the RBS units correspond to base station
sites or to base station sectors.

15. The control node in claim 12, wherein each radio resource is associated with a
frequency, a frequency range, or a frequency hopping group.
16. The control node in claim 12, wherein each of the one or more times corresponds
to a time slot or time frame.
17. The control node in claim 12, wherein the performance is determined based on
traffic load or traffic capacity using one or more of: a number of dropped calls, dropped
call rate, signal strength, signal quality, interference, delay, and an error rate.
18. The control node in claim 12, wherein the radio communications system uses
code division multiple access ,CDMA, each radio resource is associated with a code, and
each time corresponds to a transmission time interval, TTI,.
19. The control node in claim 18, wherein the electronic circuitry is configured to
instruct radio transmitters in the set of service areas to transmit at an increased power
during unblocked TTIs and at decreased power during blocked TTIs.
20. The control node in claim 12, wherein the times in each of the service areas in the
set are synchronized.
21. The control node in claim 12, wherein the time units in each of the service areas
in the set are unsynchronized, the electronic circuitry is further configured to:
determine which of the radio resource times for each of the service areas in the
set are most closely aligned,
wherein the one or more blocked times includes one or more of the determined
times.
22. The control node in claim 12, wherein the electronic circuitry is further configured
to:

compensate the desired test parameter or the determined performance for a lack of
synchronization between times in each of the service areas.
23. The control node in claim 12, wherein the electronic circuitry is further configured to:
increasing a load on the set of service areas during performance determination in
addition to blocking one or more radio resource times.
24. The control node in claim 12, wherein the electronic circuitry is further configured to:
determine a number of times not to use that avoids or limits congestion for active
subscribers below a predetermined level.

ABSTRACT

A METHOD AND APPARATUS FOR TESTING A RADIO NETWORK UNDER
INCREASED TRAFFIC LOAD
In a radio communication system that includes multiple service areas (SAs), each
service area (SA) is associated with a predetermined number or amount of radio
communication resources. Each radio communication resource can be used for a prescribed
period of time corresponding to one or more time units. An operator or other entity
identifies (S1) a set of service areas to be tested at an increased effective load. A desired
test level (e.g., increased traffic load) is determined (S2), and a corresponding number of
radio resource time units is determined to achieve the desired test level conditions (S3 and
S4). Use of radio communication resources in the determined one or more time units is
blocked (S5) for a test period. Performance by each service area during the test period is
determined (S6).

A METHOD AND APPARATUS FOR TESTING A RADIO NETWORK UNDER INCREASED TRAFFIC LOAD

Documents:

Inventors:

PCT Conventions: