2G/3G INTER-RAT HANDOVER PERFORMANCE ANALYSIS

A. Mohammed , H. Kamal2, S. AbdelWahab2

'Department ofSignal Processing, Blekinge Institute of Technology, 372 25 Ronneby, Sweden2Alcatel-Lucent, Egypt

Keywords: GSM, UMTS, Handover Performance, HandoverAlgorithms.

Abstract

One of the most interesting UMTS networks features is theirintegration with the 2G networks that provides seamless End-to-End services. Current widely deployed 2G and 3Gnetworks provide InterRAT (Inter Radio Access Technology)mechanisms enabling interoperability between them in amanner that is almost transparent to the subscriber. Theseenable maximum benefit from 3G services while ensuringwireless coverage continuity in geographically extendednetworks, where 3G coverage halls exist. These mechanismsinclude InterRAT idle mode reselection, InterRAT dedicatedmode reselection for PS (Packet Switching) services, andInterRAT handover for voice calls. The aim of this paper is toevaluate the InterRAT hand-over (HO) performance between2G and 3G networks. Analysis of the performance would bepresented in the two HO directions. We propose an analyticalmodel where the introduction of the 3G cell's load as one ofthe HO initiation parameters is considered. In addition, anInter-RAT HO ping-pong defensive mechanism is proposed.

1 IntroductionThe purpose of the inter-RAT HO (handover) procedure is tokeep the service provided to the UE (User Equipment) whilemoving away from the coverage are of one RAT (GSM orUMTS) to another (UMTS or GSM). Before the handoverprocess is completely achieved, different phases should takeplace; HO triggering conditions, measurements done by themobile UE on the neighbor cells, and selection of the bestcandidate cell that fulfills the handover criteria condition. Thehandover initiation could be based on several criteria; whethersignal level, and UE distance away from the BS, or evenbased on the desired service. Over facilitating the HOprocedures might cause the mobile terminals to handovertheir calls from one cell to another and return back to the firstcell after a very short period causing the increase of thesignaling messages on the network, which is known by thePing-Pong effect.

In the opposite sense, handovers might be designed in a way

not to be performed easily, which would increase the calldrop rate due to signal level degradation received by the UE.A tradeoff between fast and easy HO and between delayedHO exists.

In this paper we propose an analytical model where the signalquality in the UL (uplink) is also considered as one of thehandover initiation parameters for the 3G-2G HO direction.Real data from live networks is also used to complement theanalytical results.

The two handover directions; 3G-to-2G inter-RAT HO and2G-to-3G inter-RAT HO will be considered.

1) 3G-to-2G Inter-RATHO direction.

Due to the channel fading conditions faced by the UE,received signal sees fluctuations that might cause unnecessaryhandovers, even for non-moving mobiles. In order to reducethese unnecessary handovers, a proper design for thehandover initiation conditions is required. Unlike in 2Gtechnology, in 3G networks, the traffic is considerable, as ithas a direct effect on the network QoS. The noise levelincreases with the increase of active users in the cell. In thispaper we propose the consideration of the 3G-cell traffic,which is related to the noise level seen at the Node_B, for the3G-to-2G HO triggering conditions, beside the signal qualityevolution done by the UE, which is specified by 3GPPstandard. 3G-cell traffic could be evaluated by the Node B,based on the noise level received from each UE in the ULdirection. 3G-cell capacity is usually limited by the ULinterference level generated by the increased number of usersmessaging the Node B at the same time and on the samefrequency. Beside the signal quality measured by the UE andexpressed by the CPICH Ec/Jo, corresponding BER valuecould be obtained from individual link level simulations, thecell load is also added to the HO triggering conditions. TheUE signal quality reflects only the interference level receivedby the UE and not the interference level received by theNode-B. After the completion of the handover triggeringconditions, the UE goes into CM (Compressed Mode) statuswhere it could make measurements on the 2G neighbor cells.The UE sends the measurements results to the network todecide which cell is the best for HO after the fulfillment ofthe HO criteria.

The HO criterion for the Inter-RAT case is the (event 3a)criteria as specified by 3GPP standard. The criteria states that,"The estimated quality of the currently used 3G frequency isbelow a certain threshold and the estimated quality of the 2Gsystem is above a certain threshold".

This introduction of the 3G-cell traffic is expected to reducethe noise generated in the 3G cell and hence, the enhancementof the 3G system performance is anticipated.

Different from [1], the signal quality (BER) is considered asone of the HO initiation conditions in our analysis beside theUL noise level seen by the Node B. In the next section ourproposed analytical model will be presented.

2) 2G-to-3G Inter-RATHO direction.

In the GSM to UMTS direction, it's desirable to have allMobile Station (MS) terminals that are UMTS ready to beserved in the UMTS service layer. Nowadays operatorsdeploy 3G services in two strategies. The first strategy is toextend the coverage of the deployed GSM network, and theother is to co-locate UMTS sites side-by-side of the existingGSM sites, which aims at introducing new 3G services as anew service layer.

Decision to handover the UE/MS from the serving GSM cellto one possible target 3G cell must ensure acceptable radioconditions in the target 3G cell. Radio conditions of the target3G cells is defined by two main quantities [3] CPICH Ec/Joand CPICH RSCP, which are the received energy per chip ofthe primary common pilot channel of neighbor 3G cell andthe Received Signal Code Power of the primary commonpilot channel, respectively. CPICH RSCP reflects 3G radiopropagation conditions at the UE/MS location. CPICH Ec/Joreflects 3G cell signal quality at the UE/MS location, as wellas load situation in target 3G cell [4], recalling that noise levelin 3G is affected by the number ofUE/MS served by the cell.

Joint optimization of GSM to UMTS handover and UMTS toGSM handover is necessary in order to minimize call dropprobability encountered in both radio access technology andto minimize probability of unnecessary handovers. Theseunnecessary handovers result in unnecessary increase insignaling load, and is considered the main cause of theundesired ping-pong phenomena. In this paper we propose adefensive mechanism to reduce the effect of ping-pongphenomena between GSM and UMTS without a great effortin optimizing handover detection thresholds. Such defensivemechanism could be implemented in either BSS, or UTRAN.The rest of this paper is organized as follows. In Section 2 wedemonstrate handover process in both GSM to UMTSdirection and visa versa. We underpin the model byperformance metrics proposition that can be used for thepurpose of handover process optimization. Section 3 includesnumerical results extracted from field measurements. Section4 proposes the ping-pong defensive mechanism. Finally,Section 5 concludes the paper.

2 Analytical Model

performed. In this section we present the HO initiation andexecution analytical model that will be used in this paper.

Two different models will be used for the two different HOdirections: 3G-to-2G HO and 2G-to-3G HO.

1) 3G-to-2G Inter-RATHO direction.

The mobile UE performs measurements on the 2G neighborcells, once the triggering conditions are fulfilled. Thesetriggering conditions take into consideration the signal qualityat the mobile UE denoted by Qused, presented by the measuredCPICH Ec/Jo value, and also the cell load presented by theUL interference level seen by the Node B denoted by ULin-.The best cell among the neighbor cells that fulfills the HOcriteria (event 3a) is to be selected for HO execution.

Event 3a criteria could be denoted mathematically as:

QUsed< ThUsed -H3a and Qtarget +CfOtarget > Thtarget +±H3a

where Qused is the quality of the 3G cell signal measured bythe UE and presented by CPICH Ec/Jo, Qtarget is the quality ofthe target 2G cell presented by the received signal level indBm, CIO is the Cell Individual Offset of the target 2G cell.,H3a is the hysteresis margin for event 3a, Thused is thethreshold value for the received CPICH Ec/Jo, Thtarget is the2G cell received signal level threshold for event 3a.

The measured CPICH Ec/Jo reflects the signal qualityreceived by the mobile UE. The common pilot channel(CPICH) is the channel that has to be detected by all themobile UEs everywhere in the cell. If the UE is not able todetect the CPICH of certain (x) cell, this UE is said to be: "notcovered' by this (x) cell. Hence CPICH defines the coverage

area for the cell and it is transmitted by 10% of the totalNode_B power.

We propose the division of the CPICH Ec/Jo range into threeregions defined by two thresholds Th1 and Th2; a) HO isneeded region, where CPICH Ec/Jo < Th1. b) No HO isneeded region, where CPICH Ec/Jo > Th2. and c) HO mightbe needed region, where Th1 < CPICH Ec/Jo < Th2. In thethird region the UE receives somehow a degraded signalquality due to the quite increased interference level. At thispoint another metric needs to be confirmed before activatingthe CM and triggering the measurements on the neighbor 2Gcells. This metric is the UL interference level (ULint) whichpresents the 3G cell load. Figure 1 shows the three regions on

the CPICH Ec/ Jo axis.

Th Th2 CPICH-EcIo

HO might be needed

HITs nee4d No HOis7needed

Figure 1: CPICH Ec/Jo regions.The HO process passes by phases before it is completely

The handover procedures of initiation and execution for 3G-to-2G HO direction are explained in Figure 2.

yes

HO execution|

Figure 2: 3G-to-2G Handover initiation and executionanalytical model.

From the previous analytical model, conditions for a HO from3G cell to 2G cell to be occurred are;

{[Qused < Th1] or [(Th1 < Qused < Th2) and (ULint > Ith)]} and{Qused < Thused -H3a} and {Qtarget + ClOtarget > Thtarget + H3a}and {Qtarget = max [Qt, t =1,.N]} (1)

Where ULint is the UL interference level seen by the NodeB,and Ith is the interference threshold value.

2) 2G-to-3G Inter-RATHO direction

For GSM to UMTS handover the UE/ME has to measure theUMTS neighbour cells. UMTS cells that should be measuredby the UE/MS are identified and sent to the mobile onSignaling Associated Control Channel (SACCH).The networkcontrols the measurements of UMTS cells by the parameterQsearch_C sent on SACCH. Qsearch_C defines a threshold

and also indicates whether these tasks shall be performedwhen RXLEV (Received Signal level from GSM cell in dBm)of the serving cell is below or above the threshold.

Qsearch C can take the value from two different sets. Itcontrols either if the UE/MS search for 3G cells if signal levelbelow threshold (0-7): -98, -94, ..., -74 dBm, oo (always) orabove threshold (8-15): -78, -74, ..., -54 dBm, oo (never) [2].Appropriate setting of Qsearch C reduces number ofunnecessary measurement reporting and hence affectshandover probability.

Moreover, setting of this parameter directly reflects operatorUMTS deployment strategy. Setting Qsearch C such thatUMTS measurement is triggered only ifGSM level is below acertain threshold is a preferable scenario in case UMTS cellsare deployed in areas where GSM coverage holes exist. Inthis case, it becomes reasonable to set Qsearch C in 0 to 6, ashandover becomes necessary only the UE/MS experienceddegradation in GSM signal level. In case UMTS sites are co-located with GSM sites, it is reasonable to trigger UMTSmeasurements if GSM level is above a certain threshold. Thisis because in this case the GSM signal strength will be almostalways higher than UMTS signal strength. Hence it makes nosense just to wait until the GSM signal strength is degraded,as 3G signal won't be good anyway. Recommended setting is"Always search for 3G neighbors".

The UE/MS measures the neighbor UMTS cells indicated bythe network. FDD REP_QUANT parameter controls whichmeasurement quantity (RSCP or Ec/Jo) the UE/MS shallmeasure. In the rest of this paper we assume that the UE/MSmeasures CPICH Ec/Jo. In case Transmit diversity is appliedon primary CPICH the received energy per chip (Ec) fromeach antenna shall be separately measured and summedtogether into a total received energy per chip on the PrimaryCPICH, before calculating the Ec/Jo. The UE/MS reports bestUMTS cells, which is part of the neighbor cell list. Thenumber of reported UMTS cells is controlled by the GSMnetwork according to the value of the parametersFDD_MULTIRAT REPORTING in case ofUMTS FDD andby TDD MULTIRAT REPORTING in case ofUMTS TDD.

The network keeps a bookkeeping of the reported UMTScells and evaluates if handover is needed to be performedtowards a target cell. Cells are ranked for handover conditionevaluation according to the reported value of Ec/Jo, so thatbest cells are considered first. In order to compensate fastfluctuations in radio environment the network performs asliding window averaging algorithm on the reportedmeasurement quantity (CPICH Ec/Jo). TEC/IO is the handoverdetection threshold. Handover to a UMTS cell is performed ifthe condition in Equation (2) is fulfilled.

Average (Ec /IO )> TE /Io (2)

3 Results

In this section we would present how the probability that a

connection executes a handover from 3G cell to 2G cell hasbeen calculated. The calculations are based on both theanalytical model presented previously and also on themeasurements results from real live networks.

A. UMTS to GSMHandover1) Handover Probability

Figure 3 shows the CDF (Cumulative Distribution Function)for the received CPICH Ec/Jo, based on measurements in a

typical urban outdoor mobile environment. Normally the CDFshould change for different environments and also dependingon the number of installed base stations.

17lul:::: i±wI(Eb N,)*R*v (3)

where ilj, is the load factor, W is the chip rate, R is the bit

rate of the service, Eb /NO is the required energy per bit overthe noise power spectral density to establish a connectionwith a service bit rate R, and activity factor v.

Figure 5 shows the load factor for three different services (bitrates), Speech AMR (Adaptive Multi Rate) 12.2 kbps, PS64kbps, and PS384 kbps, respectively. It is clear from the figurethat the load increases with the required bit rate for thedifferent services. As the load increases, also the noise levelin the cell increases. They are directly related by thefollowing equation:

NoiseRise= 1/1 - '71u (4)CDF

Load Factor

25%

20%

15%

10%

-25 -23 -21 -19 -17 -15 -13 -11 -9 -7 -5 -3CPICH EcIIO (dB)

5%

0%12200

Se rv

Figure 3: CDF for outdoor measurements on CPICH Ec/Jo.

Figure 4 shows the CDF for the obtained Qu,sed (received

signal level) in a typical outdoor rural 2G mobileenvironment.

CDF

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

-120 -110 -100 -90 -80 -70 -60 -50 -40 -30 -20

2G Rx Level (dBm)

-10 0

Figure 4: CDF for outdoor measurements on the 2G receivedsignal level.

In [6], it has been shown that the uplink load factor caused byone connection in the WCDMA system depends on theservice the connection is establishing with the network.

Figure 5: Load factor for different services per one

connection.

Figure 6 shows the PDF of the UL interference in different3G cells deployed in a rural radio propagation environment.We can notice that the cells are facing low UL interferencelevel. These curves are the output of averaging process over

several days. Actually in nowadays 3G networks the traffic islow especially for PS services. One of these curves has beenused to calculate the HO probability in case the load exceedsa certain threshold.

PDF

0.8|-Cell 1 0.7

Cell 2 0.6

Cell 3Cell 4 0.5

+ Cell 5 0.4

Cell 6 0.3

- Cell 7 0.2

-Cell8 0.1

-94 -92 -90-108 -106 -104 -102 -100 -98 -96UL interference (dBm)

Figure 6: The PDF of the UL interference level received atthe Node B.

64000Iice Rate (bps)

384000

For loaded cells these curves are expected to shift to the right.In addition, as the UL interference level increases this mightcause call drops due to: 1) PRACH (Packet Random AccessChannel) failure 2) RRC (Radio Resources Control)connection failure. One of the major interests in deploying 3Gnetworks is to reduce the interference.

Using the previous analytical model, the cumulativedistribution functions obtained from the live networkmeasurements, beside the CDF of the chosen cell with the ULinterference values: the probability that the handoverconditions are satisfied is calculated. Figure 7 shows theprobability curves as a function of the threshold (Th1), and fordifferent values for the hysteresis margin H3a that varies from0 to 4. The other parameters have been set as shown inTable 1.

Parameter ValueCIO 0Thused -12 dBThTarget -95 dBmTh2 -5 dBULint -100 dBm

received CPICH Ec/Jo (CPICH energy per chip over the totalnoise received in the band) is less than -19 dB at the UElevel. The CPICH Ec/Jo is directly related to the CPICHEb/No by the processing gain of the established service by theuser.

This minimum CPICH Ec/Jo value is then depending on theservice; Table 2 shows the minimum CPICH Eb/No values,corresponding to -19 dB CPICH Ec/Jo value, for differentservices.

Service Minimum CPICHEb/NO

Speech AMR 12.2 kbps -44 dBCS 64 kbps -36.8 dBPS 128 kbps -33.8 dBPS 384 kbps -29 dB

Table 2: Minimum CPICH Eb/No values corresponding tominimum CPICH Ec/Jo.

In order to calculate the CDP (Call Drop Probability) due tounsuccessful HO to a 2G cell, the following model has beenassumed for a call drop to occur;

Table 1: Parameters used for 3G-2G HO probabilitycalculations.

a- The received CPICH Ec/Jo by the UE is below a certainthreshold, andb- The HO to a 2G cell is unsuccessful.

P_2G3G

- H3a=0H3a= 1H3a=2H3a=3H3a=4

0.03

0.027

0.024

0.021

0.018

0.015

0.012

0.009

0.006

0.003

-26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6Thl (dB)

-4 -2 0

Figure 7: Probability of 3G to 2G HO curves.

We can notice that, when the "HO might be needed" regionbecomes smaller, whether by increasing Th1 or by decreasingTh2, the probability that a handover occurs from the 3G cell tothe 2G cell increases. This could be explained, as the HOtriggering conditions becomes harder, the HO probabilitydecreases.

2) Call Drop Probability

The call drop rate is an important metric used to evaluate theQoS in cellular networks. Call drop could be caused by uplinkor downlink break [5]. The Downlink break is our point ofinterest in this paper. The link is assumed to break when the

This model could be expressed by the following equation;

CDP = P{(Ec Io) < ThCd} * {1- P(3G_2G)}

where Th,d is the threshold for minimum received CPICHEc/Jo value before the call drops, and P(3G 2G) is theprobability that a HO occurs from 3G cell to 2G cell.

CDP

-.H3a =01 Th =-10 0.0250.02-

0.015-

0.01

0.005-

0-20 -19 -18 -17 -16 -15 -14 -13 -12

Call Drop Ec/lo threshold (dB)

Figure 8: Call drop probability due to unsuccessful UMTS toGSM Handover.

For considerable values for the DL interference the CDPincreases. The DL interference is shown in the previousequation in the term:

P{(Ecl Io) < ThCd }

n

Using the same parameters mentioned in Table (1), with H3aset to zero, and Th1 to -OdB, the CDP has been calculatedand the results are shown inFigure 98.

Although in nowadays live 3G networks, the traffic (load) islow, but in case one PS 128 kbps user performs HO to a 2Gcell; that would decrease the cell load by about 15-16% andhence the UL interference level would be decreased by0.75dB. For a PS 384 kbps user a decrease of 1.5dB couldbe achieved.

B. GSM to UMTS Handover1) Call Interruption Time

A measurement campaign was carried out to assess callinterruption time due to handover. The test was designed suchthat several handover types are experienced by the UE/MS.During the test the UE/MS experienced 20 handovers fromGSM to UMTS, 14 handover from UMTS to GSM, and 46Intra-GSM handovers. The call interruption time due tohandover is computed as the time difference the UE/MSreceives "Handover Command" message from the old celland sends "Handover Complete" message to the new cell.This does not take into account the Handover decision time,as during that period the UE/MS is still in call.

Figure 9 illustrates the call interruption time encountered dueto Inter-RAT Handover and compared to Intra-GSMHandover. Similar performance was encountered for GSM toUMTS handover and for UMTS to GSM Handover. This isexplained by the fact that signaling path is the same for theInter-RAT Handover case. On the average the Userexperiences 200 milliseconds of call interruption, whichshould not affect voice quality compared to Intra-RAThandover.

350.00

' 300.00E

250.00.2Q 200.00

.$ 150.00

i, 100.00

G) 50.00-

0.002G-2G HO 2G-3G HO

Handover Direction

3G-2G HO

* Average* Maximum

r StdDiv

which is a typical value in case of Inter-BSC Handover.2) Handover Probability

GSM to UMTS handover probability indicates the degree bywhich the GSM network is off loaded and UMTS service ispreferred. Figure 10 shows evolution of GSM to UMTSHandover probability with different values of TEC/IO. Theeffect of setting Qsearch C is also shown. For simplicity, theeffect of averaging sliding window is not taken into account.However this should not affect the results because its rule isto compensate fast fluctuations in measured value, and notaffecting the Handover detection process. The Handoverprobability is computed numerically as:

JP(Ec Jlo> TECI )x P(RxLev < Q), Q(belo-w)HO lP(Ec Jo> TECIO)x P(RxLev> Q), Q(above)J (5)

Where Ec/Io is the CPICH Ec/Io measured by the UE/MS inGSM dedicated mode, TEC/IO is the Handover detectionthreshold based upon CPICH Ec/Jo value, RxLev is the GSMsignal level in dBm, and Q is the Qsearch_C parameter asspecified in [2][2].

Qsearch C limits probability of GSM to UMTS handover, asstated in previous section, and it should reflect Operatorstrategy of employing UMTS technology. In case UMTS isemployed for coverage extension, it is reasonable to triggerUMTS search in case GSM RXLEV is below Qsearch Cvalue. When UMTS is co-located with GSM, it is reasonableto trigger UMTS search in case GSM RXLEV is aboveQsearch_C value.

3) Call Drop Probability

Hardening Handover decision by increasing handoverdetection threshold is a good choice to reduce the chance ofunnecessary handovers. Reducing unnecessary handovers hasa direct impact on reducing Call drop in UMTS side as theprobability of underestimating radio conditions in UMTSvanishes. Also it has an influence on reducing ping-pongeffect.

t._n

0co

n

0IL

Figure 9: Call Interruption time in milliseconds for the threecases of GSM-GSM HO, GSM-UMTS HO, and UMTS-GSMHO.

0.970.8

05,-

0.4 -

0.3 -

0.2-u.10-

llN"D .K$~~l~ ~,~'b lb 'l0I,Z

Alwayes Search UMTS0.....a...Q=-2(below)

Q=-78(below)Q=-74(below)Q=-78(above)0Q=-74(above)Q=-70(above)

TEcIIo

The Intra-GSM Handovers in Figure 9 includes both Inter andIntra BSC Handovers. The maximum call interruption periodin case of Intra-GSM Handover was around 330 milliseconds,

Figure 10: GSM to UMTS Handover probability. Effect ofsetting Qsearch_C on Handover probability is shown.

A trade off between hardening and easing handover decisionis required. As Handover decision gets harder, Probability ofcall drop increases as a result of signal strength degradation inthe serving cell. Figure 11 demonstrates increase of call dropprobability in correspondence to increasing TEC/Io- Call dropprobability is simplified as [1]:

PDrop = P(RxLev < RxLevMin)x [1- PHO] (6)

where PDrop is the probability of call drop due to GSM toUMTS Handover, RxLevMin is the minimum level in dBm ofaccessing a GSM cell specified in [2] asRXLEV ACCESS-MIN, and PHO is the GSM to UMTShandover probability in Equation (5).

100 006

0.90+ GSM UMTS HO Prob Call Drop Prob|

o 80

670 0.04160 c

G40 02

0.0 2003 0.0

00

0.00 . L0.00-24 -22 -20 -18 -16 -14 -12 -10 -8 -6

TEcIlo

Figure 11: Call drop probability due to delayed GSM toUMTS Handover versus GSM to UMTS HandoverProbability.

In practice Call drop probability experienced by the UE/MS ismuch less than the values shown in Figure 11; as the figureshows only call drop that might be encountered if GSM toUMTS was not probably triggered. Typically this is not thecase, as the UE/MS has a chance to handover to another GSMcell if degradation is encountered in the serving cell. Calldrop model that capture all GSM system dynamics shouldtake into account the following Call drop events:

1) Call drop due to radio interface failure.

2) Call drop during handover execution, which include,Intra-cell, Inter-cell, Inter-BSC, and Inter-RATHandovers.

3) Call drop due to System failures, which include,Hardware failures, Software failures, andTransmission subsystem failures.

4) Call drop due to explicit preemption. As it'scommon to have a grade of service assigned toindividual subscribers, The GSM network can

explicitly pre-empt calls in case of congestion so thatit gives priority to a higher grade subscribers.

4 Inter-RAT Handover Ping-Pong DefensiveMechanism Proposal

GSM to UMTS handover as described in Section 2 isconsidered a better cell handover, in which the network willalways keep trying to send the UE/MS to the UMTS cell aslong as the radio conditions are good. This raises the risk ofping-pong effect, as the UE/MS oscillate back and forthbetween the GSM and UMTS cells in case of variations inradio conditions, and sensitive reactivity of the handovertriggering thresholds.

One possible way to minimize such undesired effect is toharden or prevent handover decision from a serving GSM cellto a target UMTS cell, if the GSM access request indicateshandover from UTRAN. We modify handover detection inEquation (2) to be:

Average (Ec Io) > TEClIO + Hp (7)

Where Hp is a penalty in dB added to the TEC/IO threshold, ifthe call is handover from UMTS.

Handover to UMTS must be enabled again by neglecting Hpvalue after a certain time. We define T1p as time during whichhandover decision is based on (4). After expiry of TIp timer thehandover decision must be based on (2) again to enable GSMto UMTS handover again. A similar defensive mechanismcould be employed in UMTS to prevent handover to GSM ifcall access request is indicating handover from GSM.

5 Conclusions

2G/3G InterRAT handover is an importanttelecommunication feature in nowadays heterogeneouswireless networks. We provided a brief description of thisfeature in the two directions, as well as analysis for itsperformance through probability calculations andmeasurements from live networks. An analytical model, andsuggested algorithms that should bring improvement to theperformance of these telecommunication procedures havebeen proposed. Tuning of different algorithms parameters waselaborated and tested in a commercial network.

6 Acknowledgements

Many people have been involved in the preparations andexecution of the tests, and have provided support led to thispaper. The authors would like to acknowledge thecontributions of their colleagues, in particular Ervin Farkas atAlcatel-Lucent Romania, Florent Colin and DidierEsclamadon at Alcatel-Lucent France, Hesham Sabry atAlcatel-Lucent Egypt, and Idris Yusof at Alcatel-LucentIndonesia.

"The views presented in this work are of the authors, anddoes not represent Alcatel-Lucent position regarding the2G/3G HO".

References

[1] W. Zhao, R. Tafazolli, B. G. Evans, "IntemetworkHandover Performance Analysis in a GSM-SatelliteIntegrated Mobile Communication System", IEEEJournal on Selected Areas in Communications, 15,(1997).

[2] 3GPP TS 45.008 "Radio subsystem link control".[3] 3GPP TS 25.215 "Physical layer - Measurements

(FDD)".[4] 3GPP TS 25.331 "Radio Resource Control (RRC)

protocol specification".[5] C. Brunner, A. Garavaglia, M. Mittal, M. Narang, J. V.

Bautista. "Inter-System Handover ParameterOptimization", QUALCOMM Incorporated, (2006).

[6] H. Holma and A. Toskala, "WCDMA FOR UMTS".Third Edition, (2004).