Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org110EFF...
Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org111com...
Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org1123. ...
Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org113res...
Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org1145. ...
Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org115Tra...
Upcoming SlideShare
Loading in …5



Published on

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org110EFFECT OF SOFT HANDOVER PARAMETERS ON CDMACELLULAR NETWORKS1N.P.SINGH , 2BRAHMJIT SINGH1Asstt Prof., Department of Electronics and Communication Engineering, National Institute of Technology,Kurukshetra - 1361192Prof., Department of of Electronics and Communication Engineering , National Institute of Technology,Kurukshetra – 136119ABSTRACTCDMA cellular networks support soft handover, which guarantees the continuity of wireless services and enhancedcommunication quality. Cellular networks performance depends upon Soft handover parameters. In this paper, we haveshown the effect of soft handover parameters on the performance of CDMA cellular networks. We consider HYST_ADD and HYST_ DROP as the Soft handover parameters. A very useful statistical measure for characterizing theperformance of CDMA cellular system is the mean active set size and soft handover region. It is shown throughnumerical results that above parameters have decisive effect on mean active set size and soft handover region andhence on the overall performance of the soft handover algorithm.Keywords: Code-Division Multiple Access(CDMA), Soft handover , Shadow fading , Soft handover region,Mean active set size1. INTRODUCTIONCode-Division Multiple Access (CDMA) basedcellular standards support soft handover, whichmakes smooth transition and enhancedcommunication quality. Soft handover offersmultiple radio links to operate in parallel. Mobileusers are separated by unique pseudo-randomsequences and multiple data flows are transmittedsimultaneously via radio interface. The UserEquipment (UE) near the cell boundary isconnected with more than one Base Station (BS).Consequently, in soft handover UE is able to getbenefit from macrodiversity. Soft handover can,therefore, enhance both the quality of service andthe capacity of CDMA based cellular networks [1-2].Soft handover problem has been treated in literaturefor performance evaluation of the algorithms usingboth analytical and simulation methods [3-7, 20].An analytical model has been proposed in [8] toevaluate cell assignment probabilities to computeoutage probability, macrodiversity gain, andsignaling load.Soft handover is associated with active set and itssize. The inclusion and drop of a particular BSin/from the active set is determined by the initiationtrigger utilized for soft handover algorithm.Initiation trigger include, received pilot signalstrength, Carrier to Interference ratio, Bit-Error-Rate, Energy per bit to noise power density. In thepresent work, we consider received pilot signalstrength as the initiation trigger. It is easilymeasurable quantity and directly related to thequality of the communication link between UE andBS and the most commonly used criterion forhandover initiation trigger. In the present work, wehave utilized received pilot signal strength as theinitiation trigger.Due to the random nature of the received signal atUE, there are frequent inclusion and drop of BS(s)in the active set. Active set consists of those basestations which are connected with UE and softhandover region [15] is that region in which UE isconnected with more than one base station. It isessential for properly designed soft handoveralgorithm [16, 17, 18, 19, 21, 22, and 23] to reducethe switching load of the system while maintainingthe quality of service (QoS). In this paper, we haveconsidered mean active set size and soft handoverregion as the metric for performance evaluation ofthe soft handover algorithm. It is directly related toperformance of handover process and is required tobe minimized.The rest of the paper is organized as follows.Section 2 describes the system model used forTempus Telcosys
  2. 2. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org111computer simulation; Soft handover algorithmfollows cellular layout and radio propagationassumptions. It is followed by description of softhandover algorithm in section 3. Numerical resultsare obtained, plotted and discussed in section 4.Finally, conclusion and future work are drawn insection 5 and 6 respectively.2. THE SYSTEM MODELWe consider a simple cellular network consisting oftwo representative cells supported by two BS(s).Both of the BS(s) are assumed to be located in thecenter of the respective cell and operating at theequal transmitting power as depicted in Fig. 1.Hexagonal geometry of the cell has beenconsidered. The UE moves from one cell to anotheralong a straight line trajectory with constant speed.BSs are assumed to be D meters apart.Fig. 1 Cellular configurationReceived Signal Strength (RSS) at UE is affectedby three components as follows:(i) Path loss attenuation with respect to distance(ii) Shadow fading(iii) Fast fadingPath loss is the deterministic component of RSS,which can be evaluated by outdoor propagationpath loss models [9-10]. Shadowing is caused dueto the obstruction of the line of sight path betweentransmitter and receiver by buildings, hills, treesand foliage. Multipath fading is due to multipathreflection of a transmitted wave by objects such ashouses, buildings, other man made structures, ornatural objects such as forests surrounding the UE.It is neglected for handover initiation trigger due toits short correlation distance relative to that ofshadow fading.The UE measures RSS from each BS. Themeasured value of RSS (in dB) is the sum of twoterms, one due to path loss and the other due tolognormal shadow fading. The propagationattenuation is generally modeled as the product ofthe thη power of distance and a log normalcomponent representing shadow fading losses [9].These represent slowly varying variations even forusers in motion and apply to both reverse andforward links. For UE at a distance ‘d’ from BS1,attenuation is proportional to1 0( , ) 1 0d dζηα ζ = (1)where ζ is the dB attenuation due to shadowing,with zero mean and standard deviationσ .Alternatively, the losses in dB are( , )[ ] 10 logd dB dα ζ η ζ= + (2)where η is path loss exponent. The autocorrelationfunction between two adjacent shadow fadingsamples is described by a negative exponentialfunction as given in [11]. The measurements areaveraged using a rectangular averaging window toalleviate the effect of shadow fading according tothe following formula [12]. Let di denote thedistance between the UE and BSi, i=1, 2.Therefore, if the transmitted power of BS is Pt, thesignal strength from BSi, denoted Si,(d) i=1, 2, canbe written as [14].Si(d)= Pt - ( , )idα ζ (3)101ˆ ( ) ( )Ni i nnwS k S k n WN−== −∑ i = 1, 2 (4)where; ˆiS is the averaged signal strength and iS isthe signal strength before averaging process. Wn isthe weight assigned to the sample taken at the endof (k − n) thinterval. N is the number of samples inthe averaging windowNw =10NnnW−=∑ .In the case of rectangular window Wn = 1 for all n.The size of averaging window should be selectedjudiciously as larger size of the window may notdetect the need of handover initiation trigger atright time. On the other hand, smaller window maycause ping-pong effect, which is not desirable. Inthis work, first of all we have determined theoptimum value of the size of averaging window forgiven system parameters and then the effect ofpropagation parameters is analyzed for that typicalsetting of averaging window size.Shadow fading in the present work is modeled asfollows [13]2( ) ( 1) (1 ) (0,1)ik k Wζ ρζ σ ρ= − + − (5)where ρ is correlation coefficient and (0,1)Wrepresents truncated normal random variable.BS1BS2Cell1UECell2Tempus Telcosys
  3. 3. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org1123. SOFT HANDOVER ALGORITHMCDMA systems support the soft handover process.Active set is defined as the set of base stations towhich the UE is simultaneously connected. In softhandover region, UE is connected with more thanone base station. A soft handover algorithm isperformed to maintain the user’s active set. For thedescription of the algorithm, the followingparameters are needed.-HYST_ADD: Hysteresis for adding.-HYST_DROP: Hysteresis for dropping.Fig.2 Soft handover algorithmActive set is updated in following manner:- If active set consists of BS1 and 1 ( )S d∧> 2 ( )S d∧and ( 1 ( )S d∧- 2 ( )S d∧) is greater than HYST_ADD,Active set consists of BS1- If |( 1 ( )S d∧- 2 ( )S d∧)| < HYST_ ADD, Active setconsist of BS1 and BS2 .- If active set consists of BS1 and BS2 and |( 1 ( )S d∧-2 ( )S d∧)| becomes greater than or equal to HYST_DROP, Active set consist of that base station whoseaverage signal strength is higher. Base station withsmaller signal strength will be dropped from theactive set.Where 1 ( )S d∧, 2 ( )S d∧is the averaged signalstrengths from BS1 and BS2 respectively.4. RESULTS AND DISCUSSIONIn this section, mean active set size and softhandover region has been computed as the functionof different system parameters and characteristicparameters of radio propagation environment.Numerical results for this performance metric areobtained via computer simulation for the systemparameters indicated in Table 1. System simulationis performed in Matlab 7.0. To obtain mean activeset size and soft handover region for each systemparameter setting, 5000 runs of the simulationprogram are performed.Table 1.System parameters for system simulationD = 2000 m Distance between twoadjacent BSsη =4.0 Path loss exponentds = 1m Sampling distancePt = 0.0dBm Base station transmitterpowerρ =0.9512 Correlation coefficientσ =8.0 Standard deviationN=20 Averaging window sizeFig. 3a & 3b show the effect of HYST_ADD on themean active set size. Mean active set size increaseswith the increase of the add hysteresisHYST_ADD. Mean active set size is maximum atthe boundary and decreases as UE moves towardBS. High mean active set size increases signalingload on network and very low value of mean activeset size may not give full advantage of macrodiversity but it will increase the probability ofoutage.Fig. 4a & 4b shows the variation of mean active setsize with distance for different value of drophysteresis HYST_DROP. Mean active set sizeincreases as drop hysteresis HYST_DROP isincreased. A large increase in mean active set sizeis observed. The reason for this result is attributedto the fact that greater value of drop hysteresis willincrease active set size for larger area and hencesoft handover region will increase. Large softhandover region will increase interference and thisis not desirable. Therefore drop hysteresis shouldneither be very large nor very small.Finally, Fig. 5 and Fig. 6 depicts the variation ofsoft handover region with add hysteresis and drophysteresis. This plot shows that, as add hysteresisand drop hysteresis are increased, soft handoverregion increases rapidly. Large soft handoverregion increases interference and hence capacitydecreases. Very small soft handover region mayincrease probability of outage. This is an interestingBSPowerreceivedat UEDistanceBS1+ BS2BS1 BS22( )S d∧HYST_ADDHYST_DROP2 ( )S d∧Tempus Telcosys
  4. 4. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org113result and should be taken into consideration whiledesigning soft handover algorithm0 200 400 600 800 1000 1200 1400 1600 1800 200011. 3a Effect of HYST_ADD on mean active setsize for given HYST_DROP=14dBm.2 4 6 8 10 12 3b Effect of HYST_ADD on mean active setsize for given HYST_DROP=14dBm.0 200 400 600 800 1000 1200 1400 1600 1800 200011. 4a Effect of HYST_DROP on mean active setsize for given HYST_ADD=1dBm.4 6 8 10 12 14 4b Effect of HYST_DROP on mean active setsize for given HYST_ADD=1dBm.2 4 6 8 10 12 1424252627282930313233HYST__ADD(dBm)SofthandoverRegion(%)Fig. 5 Effect of HYST_ADD on soft handoverregion for given HYST_DROP=14dBm4 6 8 10 12 14 1651015202530HYST__DROP(dBm)SoftHandoverRegion(%)Fig. 6 Effect of HYST_DROP on soft handoverregion for given HYST_ADD=1dBm.Tempus Telcosys
  5. 5. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org1145. CONCLUSIONIn this paper, we have tried to understand theimpact of soft handover parameters HYST_ADDand HYST_DROP on soft handover performance,which was measured in terms of mean active setsize and soft handover region. Simulation resultsshow that soft handover parameter setting has adecisive effect on the handover performance. Bycareful tuning and setting appropriate values for thehandover parameters, a higher system performancecan be achieved6. FUTURE WORKThere is a need to integrate WLAN, WMAN, andCellular networks in a best possible way. WLANgenerally provide higher speeds and morebandwidth, while cellular technologies generallyprovide more ubiquitous coverage. Thus the usermight want to use a WLAN connection wheneverone is available, and to fail over to a cellularconnection when the wireless LAN is unavailable.Vertical handover refer to handover from onetechnology to another technology to sustaincommunication. Mobile IP provides mobilityamong different technologies at network layerbut the problem is that during handover it breaksthe connection which causes some packet loss. Softhandover algorithm can be used in future with trustassisted handover approach in hybrid wirelessnetworks.REFRENCES:[1] K. S. Gilhousen, I.M. Jacobs, R. Padovani, L.Weaver, “ Increased Capacity using CDMAfor Mobile Communications”, IEEE Journalon Selected Areas in Communications, Vol. 8,May 1990, pp. 503-514.[2] A.J. Viterbi, A.M. Viterbi, K.S. Gilhousen, E.Zehavi, “Soft Handoff Extends Cell Coverageand Increases Reverse Link Capacity,” IEEEJournal on Selected Areas inCommunications, Vol. 12, October 1998, pp.1281-1287.[3] N. Jhang, J.M. Holtzman, “ Analysis of CDMASoft-Handoff Algorithm”, IEEE Transactionson Vehicular Technology, Vol. 47, May 1998,pp. 710-714.4] B. Homnan, V. Kunsriruksakul, W.Benjapolakul, “ A Comparative Evaluation ofSoft Hnadoff between S-95A and IS-95B/cdma2000”, Proc. IEEE VehicularTechnology, 2000, pp. 34-37.[5] V. Vassiliou, J. Antoniu, A. Pitsillides, G.Hadjipollas, “ Simulating Soft Handover andPower Control for Enhanced UMTS”, IEEE16thInternational Symposium on Personal,Indoor and Mobile Radio Communications,2005, pp. 1646-1651.[6] D. Zhang, G, Wei, J. Zhu, “Performance ofHard and Soft Handover for CDMASystems”, IEEE Vehicular TechnologyConference 2002, pp. 1143-1147.[7]. J. Reig, “ Capacity Analysis in DownlinkWCDMA Systems Using Soft HanodverTechniques With SIR-Based Power Controland Site Selection Diversity Transmission”,IEEE Transactions on Vehicular Technology,Vol. 55, July 2006, pp. 1362-1372.[8] A.E. Leu B.L. Mark, “ Discrete-time Analysisof Soft Handoff in CDMA CellularNetworks”, Proc. IEEE Vehicular TechnologyConference, 2002, pp. 3222-3226.[9] V. K. Garg, J.E. Wilkes, “Principles andApplications of GSM”, Prentice Hall PTR,Upper Saddle River, NJ 07458.[10]. K. Feher, “ Wireless Digital Communications,Modulation and Spread SpectrumApplications”, Prentice-Hall of India, NewDelhi-2001.[11] V. M. Gudmundson, “Correlation Model forShadow Fading in Mobile RadioCommunication System,” Electronics Letters.,vol.27, November 1991, pp.2145-2146.[12] G.E. Corazza D. Giancristofaso, F. Santucci,“Characterization of Handover Initiation inCellular Mobile Radio Networks”, IEEETechnology Conference, 1994, pp. 1869-1872.[13] ] N.P.Singh, Brahmjit Singh “Effects of SoftHandover Margin under various Radiopropagation parameters in CDMA CellularNetworks”,IEEE conference on WCSN-2007,pp 47-50[14] Brahmjit Singh “An improved handoverAlgorithm based on signal strength plusdistance for interoperability in mobile cellularnetworks” Wireless Personal Communication(2007), 43:879-887.[15] Xinbing Wang, Shiguang Xie, and Xiuwen Hu“Recursive Analysis for Soft HandoffSchemes in CDMA Cellular Systems” IEEETempus Telcosys
  6. 6. Journal of Theoretical and Applied Information Technology© 2005 - 2010 JATIT& LLS. All rights reserved.www.jatit.org115Transaction on Wireless Communications,Vol.8, NO. 3, March 2009[16] Qualcomm Corporation, “Diversity-Handovermethod and performance”, ETSI SMG2Wideband CDMA Concept GroupAlphaMeeting, Stockholm, Sweden, Sept. 1997[17] EIA/TIA/IS-95A, ‘Mobile Station-BaseStation Compatibility Standard for Dual-Mode Wideband Spread Spectrum CellularSystem’, Telecommunication IndustryAssociation, Washington CD, May 1995[18] J. Laiho-Steffens, M. Jasberg, K. Sipila, A.Wacker and A. Kangas, “Comparison of threediversity handover algorithms by usingmeasured propagation data”, proceedings ofVTC’99 Spring, Houston, TX, pp. 1370-1374,May 1999[19] Universal Mobile Telecommunications System(UMTS); Radio Resource ManagementStrategies, (3GPP TR 25.922 version 5.0.0Release 5)[20] R. P. Narrainen and F. Takawira,“Performance analysis of soft handoff inCDMA cellular networks,” VehicularTechnology, IEEE Transactions on, Vol. 50,Issue. 6, pp. 1507-1517, Nov. 2001[21] X. Yang, S. Ghaheri-Niri and R. Tafazolli,“Performance of power triggered and Ec/N0-triggered soft handover algorithms forUTRA,” 3G Mobile CommunicationTechnologies, Conference Publication, No.477, pp. 7-10, 2000[22] X. Yang, S. Ghaheri-Niri and R. Tafazolli,“Evaluation of soft handover algorithms forUMTS,” Personal, Indoor and Mobile RadioCommunications, 2000 11th IEEEInternational Symposium on, Vol. 2, pp. 772-776, 2000[23] S. S. Wang, S. Sridsha and M. Green,“Adaptive soft handoff method using mobilelocation information,” VehicularTechnologyConference, 2002, VTC Spring2002. IEEE 55th, vol. 4, pp. 1936- 1940, 2002AUTHOR PROFILES:N.P. Singh receivedB.E. & M.E. degree inElectronics andCommunicationEngineering from BirlaInstitute of Technology,Mesra Ranchi, India in1991 & 1994 respectively and presentlypursuing PhD degree in Electronics andCommunication Engineering at NationalInstitute of Technology Kurukshetra,India. The current focus of his researchinterests is on radio resource management,interworking architectures design, mobilitymanagement for next generation wirelessnetworks.Brahmjit Singhreceived B.E. degreein ElectronicsEngineering fromMalviya NationalInstitute ofTechnology, Jaipur in1988, M.E. degree in Electronics andCommunication Engineering from IndianInstitute of Technology, Roorkee in 1995and Ph.D. degree from Guru Gobind SinghIndraprastha University, Delhi in 2005(India). Currently, he is Professor,Electronics and CommunicationEngineering Department, NationalInstitute of Technology, Kurukshetra,India. He teaches post-graduate andgraduate level courses on wirelessCommunication and CDMA systems. Hisresearch interests include mobilecommunications and wireless networkswith emphasis on radio resource andmobility management. He has published65 research papers inInternational/National journals andConferences. Dr. Brahmjit Singh receivedthe Best Research Paper Award from TheInstitution of Engineers (India) in 2006.Tempus Telcosys