By Kevin Martin,2014-05-20 06:48
6 views 0
The Research of Optimization algorithms of Vertical handoff in heterogeneous network Subtitle as needed College of Information Science & Technology, Qingdao University of Science & Technology Qingdao, 266061, P.R. China e-mail: Authors Name/s per 2nd Affiliation (Author) line 1 (of Affiliation): dept. name of organization line 2: name o..

    The Research of Optimization algorithms of Vertical handoff in heterogeneous


    Subtitle as needed

College of Information Science & Technology, Qingdao Authors Name/s per 2nd Affiliation (Author) University of Science & Technology line 1 (of Affiliation): dept. name of organization Qingdao, 266061, P.R. China line 2: name of organization, acronyms acceptable e-mail: line 3: City, Country

    line 4: e-mail:

    AbstractIt is heterogeneous wireless Networks (HWN) centralized control transform to mobile terminal to select aggregation foundation for vertical handoff (VHO)and that optimum wireless network.

    The process of VHO usually includes three stages: 1) is key characters of core technology for multi-network

    aggregation. The currently mechanism handoff trigger is network discovery, that is, users detect accessible wireless analyzed, a best network selected (BNS) framework is designed, networks; 2) handoff decision, that is, according to the parameters of network terminal and user terminal of characteristics of themselves and the network performance, effecting VHO algorithm, and which also include norm of users evaluate all accessible networks and determine handoff capable parameters, classifying the user business and decision; 3) handoff performing, that is, users can perform determining the weight of parameters of network according to vertical handoff in accordance with the result of the handoff different user Business are studied. An optimization algorithm decision. The handoff decision stage is the key above three of VHO with variable weights is proposed in mobile terminal. stages, because their algorithm directly relates to the Weights of the parameters are variably adjusted according to effectiveness and reliability of VHO, and also effect QoS and different service of users, the algorithm is significant the network performances of uses. improvement compared with traditional methods to effectively

    improve the user’s satisfaction degree and the integration RESEARCH II. RELATEDperformance of the network. In order to achieve the ABC, a number of currently

    solutions are proposed for vertical handoff in heterogeneous Keywords- HWN; VHO; Norm of Parameters; Variable

    weight; Cost function network, a algorithm which just considers Received Signal

    Strength (RSS) and threshold is proposed in the document

    [1]. Differences between different networks are not reflected I. INTRODUCTION comprehensively about this algorithm. Analytic Hierarchy Being development of mobile communication technology, Process (AHP) is introduced to calculate the weight of each the wireless network of the next generation should be factor and network performance is measured combined with heterogeneous networks which contain varies of wireless Grey Relational Analysis in document [2]; networks. HWN is a network that relate to two or more than Fuzzy logic is used to select the best network in two wireless access systems to implement different wireless document [3].The algorithms of document [2] and [3] were access technologies or adopt the same wireless access some defect, e. g. existing mass of computing, required input technology but belonging to different operators in the lot of fuzzy configuration data, it seems very Inflexible. wireless network, for example, CDMA2000, WCDMA, TD-Document [4] proposed that weights of different network CDMA, WiMax, and IEEE802.a/b/g, WLAN, Bluetooth and performance parameters are determined by the method of other radio access network are all can be aggregated with artificial Intelligence. Considering that WLAN bandwidth is together in HWN model. The integration among wider and costs less than the cellular network, document [5] heterogeneous networks can be classified tight coupling and proposed the approach of vertical handover when users reach loose coupling mode. However, the wireless network of the the covered range of WLAN from cellular networks. J. Wang, next generation is widely regarded as the loose coupling et. al. proposed input weights of network parameters by user framework based on the IP core network, that is, access and to compute costs of network by using a utility function, networks access to the IP backbone network by the network and then select access network by comparing the level of the gateway. The VHO among heterogeneous networks is the network cost function [5]. The algorithm of J. Wang involves important study content of the mobile management and a large number of factors, including rates, energy operation in mobile traffic, which directly relates to the consumption of network interface, bandwidth and so on. effective utilization of wireless resources and the quality of Document [7] proposed a vertical handoff algorithm with service (Qos) to users. The VHO is wireless network variable weight, but it did not propose a method to specify

    the weights. Due to the cost function method taking into

    account many characters of the access network and the user’s network in the document [8], and then the network will terminal, it can guarantee the network's integration return the information of all the wireless networks around performance and it is the main research directions for mobile terminal back .Here are some QoS parameters of vertical handoff algorithm. networks shown in table 1.

    In currentlythe main issue of VHO algorithm based on TABLE I. NETWORKS QOS PARAMETERS cost function research are as follows: (1) There are many

    differences between different algorithms in the set of Typical Bandwidth Mobility Rates End to objective cost function and the selection of limited Parameters Support end Mbps conditions, It is difficult to comprehensive evaluate the Delayalgorithm; (2) Many VHO algorithms described VHO ms between only two networks ,it is not a optimal algorithm for WLAN 20 Low Free 80 the integration heterogeneous network; (3) The set of GPRS 0.056 High 150 0.1/min optimization cost function and the selection of limited

    WCDMA 2 High 125 conditions show that the parameters have an effect on 1/min

    algorithm performanceit is also key issue of the algorithm 802.16d 10 Low Charge 100

    and is also lacking in the current study. for example, many monthly cost function method only take on Logarithmic or index on C. Terminal State Detection various network parameters, rather than normalized

    reasonably parameters; (4) Different business of users needs Terminal state detection module obtains terminal different indicators of different network parameters, It is a dynamic information by calling the operating system's API significant problems that must be considered which need to (Application Programming Interface) and provides Qos optimize the vertical handoff algorithm for the effective use parameters which is the mobile terminal expected to decision of wireless resources, and to reflect different user needs device module. There are some real-time service QoS On the condition of loosely coupled in the heterogeneous requirements shown in table 2. Terminal state detection network, a BNS framework is designed and an optimization module also detects the current available battery capacity and algorithm of VHO based on cost function with variable other information which are transmitted to the decision weights is proposed in mobile terminal. device.


    Generally, the best network selection (Best Network Media Application Bandwidh End to end delay Selection, BNS) has two modes: terminal-based and Audio Voice 4-25kb/s Best network-based. The selection algorithm of terminal control conversation Delay<150ms is more flexible and easier to achieve, it can easily perceive Maximum wireless environment of mobile terminal and reflect the delay<400ms wish of users in a maximum degree. Because network Video Video Phone 32-Best conditions change as the user's moving, users need to select 384kb/s Delay<150ms a current best network by dynamically assessing conditions Maximum of the networks and the terminal state, so a BNS framework delay<400ms shown in figure 1 is designed in the mobile terminal. Data 1 Telemetry <28.8kb/s Required

    Two-way delay<250ms A. Advertisement Brand control Base stations of wireless networks send the current QoS Data 2 Interactive <1KB Required information to advertisement brand, advertisement brand Games, delay<250ms regularly broadcasts the current QoS information to the Remote Login public channel. So the current QOS information should be received by network QoS detection

    D. Decision Device B. Network QoS Detection

    Detection device judges every network whether it is an Network QoS parameters of the heterogeneous network accessible network using network Qos parameters collected which are sent to decision device for computing network by network QoS detection module and expected Qos priority are collected by network QoS detection. There is a parameters collected by terminal state detection module, for method which can be used to get the Qos parameters of example, decision device will shut up a certain network network: the network QoS detection module will collect the when it’s RSS is less than threshold value reported by current QoS information which is sent to the public channel physical layer. The decision device can also assess the by the advertisement brand form the public channel when the velocity of mobile terminal by the RSS and shut up a certain decision device sends inquiry request to the network QoS network when it’s velocity exceeds the current maximum detection module. For example, TS23.402 of 3GPPLTE

    proposed that inquiry request when needed is sent to wireless

    velocity which the certain network supportedand vice is the available network bandwidth of network Bifversa. ,is the total bandwidth of the network , reflects iiThe decision device also have many other functions GBiiincluding norm of the network parameters and determining the proportion of available bandwidth in the total bandwidth the weight of network parameters according to different and reflects the network bandwidth load . The document [11] users service, then the decision device gives a decision which defined the bandwidth returns function show as formula (3) is the best network.

    bUE. Decision Execution ?1,?aB?bLDecision Execution receives the VHO order coming form ?aBfsaLbU;;(;(,)~,bBaB(decision device and execute to select the best network. ULBB??0,baLB)IV. NORM OF PARAMETERS

    In order to select the best network, the network asis the system state vector;is the target network of performances of accessible networks are needed to be bVHO;is the available network bandwidth of the target; acompared according to an objective function. Because the

    heterogeneous networks have different structures and LUand are the minimum and maximum bandwidth BBcharacters, there are many differences between the value requirements for the users .The formula (3) measures the ranges of networks parameters, the evaluation of network candidate network conditions on the point of bandwidth performance and comparison of network parameters are requirements for users. The bandwidth return function difficult. In order to realize fair parameter and comparison of reflects ability of the candidate network providing bandwidth networks performance, norm of the parameters are adopted. for users excellently

    A. System available Bandwidth B. Norm of RSS

    Systems available bandwidth is an important indicator of Taking cellular network and WLAN for example, system performance. From the view of user, larger systems because cellular networks Bs and WLAN AP's transmission available bandwidth means higher transmission rates and power and coverage of two networks are very difference, the shorter transmission delay, therefore, it is one of the major BS and the AP's signal strength are also very different. considerations factor at the time of network selection. Since Generally, the coverage of WLAN is far less than cellular the maximum bandwidth value of each network and the network, the distance between the AP and the user is closer value range of available bandwidth provided for uses are than the distance between the BS and the user ,so the AP different, the program which is adopted by directly RSS is stronger than the BS PSS comparing the RSS form comparing the available bandwidth will affect the fairness the BS and the AP directly, even if the user receives the AP's and effectiveness of VHO decision, the network bandwidth RSS which is not enough strong the VHO decision still should be normalized and then compared the normalized selects to access the WLAN .In this case, it is difficult to parameters. The document [9] analysis many system guarantee communication link quality and the network QoS performance associated with the VHO takes the network in WLAN .Therefore, in order to achieve fair and effective bandwidth for example, and proposes to select candidate comparison of RSS , the RSS should be normalized based on network of the biggest normalized bandwidth as the VHO the received power threshold and maximum power values of objective network, the document [9] exists network, the ithe network. The document [10] gives the normalized RSS bandwidth of the network is, the definition of the iDiof network i shown as formula (4).

    normalized bandwidth of the network is showed as i

    formula (1). ethPPii~;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;SimaxthDiPPBii~;;;;;;;;;;;;;;;;;;;;;;;;;;;;!?;normmax(D,D,;,D)12NeThe formula (4) proposes that is the current RSS PiThe formula (1) simply normalizes the bandwidth and threalizes unified comparison with other system parameters, iwhich received form network , is the acceptance Pibut it is difficult to show the proportion of available maxbandwidth in the total bandwidth, that is the network load threshold,is maximum signal strength. The formula (4) Piconditions. The document [10] proposes the method of norm reflects the relations between currently received signal of bandwidth shown as formula (2) strength of users and the maximum signal strength, it

    excellently measures user currently received signal of

    Biifnetwork. So this method is reasonable. ~;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GiBi

    TABLE III. CLASSIFICATION OF ALL USER SERVICE AND PRIORITY C. Norm of the Network Delay LEVEL OF THE PARAMETERS Since each network adopts different resource

    Classification Typical Signal Bandwi-End Sermanagement mechanisms and user requests access

    of user service strength dth to vicmechanism, accordingly, connection delay produced by each network is different. The document [12] gives the delay service end e return function shown as (5). delay lev


     Real-time Voice 4 2 1 3

    dLclass conversation, ?1,?aD?chat with qq, Ud?DaSsaLdU(,)~,;;;;;;;;;;;;;;,? DDaD(interactive ULDD?games ?0,d?UaD)Loss of large Video 3 1 4 2

    flow class surveillance, d The formula (5) proposes that is connection delay of aweb

    browsing LandUobjective network;are separately the minimum DD

    and maximum connection delay of users which are allowed. VI. ALGORITHM AND CONCLUSION The formula (5) can effectively assess the connection delay

    The basic expression the algorithm is shown below. The of different objective network by comparing connection delay request of users and the current connection delay of function of ?1 is shutting up a certain network when the

    each network. RRS is weak while the function of ?2 is escaping form

    pingpang of VHO. V. VARIABLE WEIGHT OF VHO ALGORITHM thIf?1 Then shut up network //?1 iRSS?RSSiiThe existing vertical handoff algorithms based on cost is fixed value function generally adopts fixed weights. Different service of VVElse assess //is the velocity of the mobile terminal, users have different Qos requests for network, the priority of 00

    parameters are different because of the effect of different Vis got by assessing the RSS. 0user service for network parameters .for example, the request thth? If Then shut up network i //is the of the voice service is that the end to end delay is less than vvv0ii150ms and the bandwidth needs only 16kbs, so a larger maximum velocity which network supports iweight should be given to network delay and a smaller R,B,D,CW,W,W,W Else norm) and get iiiiRBDCweight should be given to the bandwidth. However, the // WR, WB ,WD and WC are got by user service request of browsing web business is that bandwidth needs more than 200ms, and a longer network delay is allowed, so WWRB Then // (7) S~WDWCDiCia larger weight should be given to bandwidth and a smaller RBiiweight should be given to the network delay. A VHO S Then S= min [], put out b =i ialgorithm with variable weight based on the user service is proposed, which realizes the best use of the radio resource. S~S If Then Keep Connection a // a is the aCost function of algorithm can be described as the formula Susing network. is the value of the using network a(6). computed by the formula7

    thSSSWW Else if (+?2 and ) RSSa?RSSaRBa;;;;;;;;;;;;;;;:?;S~argmin(WDWC)DiCiThen Keep Connection a RBii // ?2 is fixed value

    thSSS Else if (+?2and ) RSSa?R,B,D,CThe formula (6) proposes that are RSSaaiiiiThen VHO to b separately norm of signal strength, norm of bandwidth, Else VHO to b norm of network delay and service Level, and End W,W,W,W are separately weight of the parameters, RBDC weight values can be dynamically adjusted. Classification of A large number of VHO algorithms are analyzed in this user service and priority level of the parameters are shown in paper, a BNS framework is designed, the parameters of table 3.The number 1 represent the highest priority, and the network terminal and user terminal of effecting VHO number 4 represent the lowest priority. algorithm, and which include norm of capable parameters classifying the user service and determining the weight of

    parameters of network according to different user service are ;studied. VHO algorithm with variable weight is proposed,

    [5] ZHANG QGUO C XGUO Z Het a1Efficient Mobility the algorithm makes different requirements of users Management for Vertical Handoff between WLAN and embodied in every weight of optimization object function of WLAN[J]1EEE Comm Mag200311102108 VHO algorithms. It merges advantages of existing [6] H.J.WallgR.H.KatzJ.Giese. Policy-enabled handoffs across algorithms, and makes them improved and systematic, which heterogeneous Wireless networks in Mobile Computing Systems and can effectively satisfy users and make good use of network Applications1999. proceedings. WMCSAresources. This method is more progressive than traditional 99.SecondIEEEWOrkshoPon1999:51-06

    one.. [7] Chai rong,xiaomin.Perfomance parameter analysis and algorithm study of vertical handoff for heterogeneous networks. Chongqing University of Posts and Telecommunications University report REFERENCES (Science Report). Feb. 2010. [1] Tripathi N DReed J HVanlandinghum H FAdaptive Handoff [8] 3GPP.Arehitecture enhancements for non-3GPP accesses. Algorithm for Cellular Overlay Systems using Fuzzy TS23.402vol.VS.4.12009-01 Logic[J]1EEE 49th Vehicular Technology Conference16-20 May 199921413-141 [9] 161 NASSER NHASSWA AHASSANEIN HHandoffs in Fourth Generation Heterogeneous Networks[J]IEEE Comm Mag[2] Song QJamalipour AA Network Selection Mechanism for Next 20061196103 Generation Networks[J]IEEE ICCMay 20051418-1422 [10] SHEN WZENG Q ACostFunctionBased Network Selection [3] X.LiuJ.Ling-geH. Chen. A Novel Fuzzy Logic Vertical Handoff Strategy in Integrated Wireless and Mobile Networks[J]IEEE Algorithm with Aid of Differential Prediction and Pre-Decision Trans Vehi Tech200857(6)37783788 Communications2007 ICC07.IEEE International Conference on2007:5665-5670 [11] STEVENS Navarro ELIN Y XWONG V W SAn MDP. Based Vertical Handoff Decision Algorithm for Heterogeneous Wireless [4] G.Lee,P.Faratin,s.Bauer,et al.Wroclawski.A user-guided cognitive Networks[J]IEEE Trans Vehi Tech200857(2)1243-1254 agent for network service selection in Pervasive computing environments. in Pervasive Computing and Communications[12] SONG W.ZHUANG W HCHENG Y. Load Balancing for 2004.PerCom2004.Proeeedings of the second IEEE Annual Cellular/WLAN Integrated Networks[J]IEEE Network2007Conference on2004:219228 1:27-33

    Public Channel Advertisement Network QOS Brand Detection

    Inquiry Network QoS

    request parameters

    Terminal Decision State WLAN Device Terminal Qos Detection

    parameters CDMA2000 GPRS

    Ad hoc



    Fig.1 BNS Framework

Report this document

For any questions or suggestions please email