On Selection Policy In Heterogeneous Wireless Networks

With the rapid development of mobile communication technology, the heteroge-neous wireless networks which are suitable for different environments can providehigher quality of service for the users. The proper network selection policy which cansatisfy broadband access and seamless mobility has become a problem demandingprompt solution. In mobile communication system, the quality-of-service(QoS) andnetwork performance are influenced by cell coverage, received signal strength(RSS),delay, bandwidth, signal-to-noise ratio(SNR). In this dissertation, the selection policiesin the heterogeneous wireless networks are discussed, the main achievements andresults are listed as follows:1. In heterogeneous wireless networks, a new network selection strategy basedon bandwidth resource allocation is proposed to manage the problem caused by thelimited bandwidth resource. The proposed strategy focuses on maximizing the totaltransmission rate and the bandwidth utilization with the bandwidth resource constraint.The transmission rate is formulated as an object function with the constraint of band-width resource. An optimal solution to bandwidth resource allocation is obtained, andthen the user is assigned to the proper network. Both the mathematical analysisand simulations show that the proposed network selection strategy can obtain higherthroughput and resource utilization.2. Combining delay constraint with instantaneous transmission rate, a new net-work selection policy is proposed for heterogeneous wireless communication systemsusing effective capacity. In reality, different services have different delay constraints,while the traditional Shannon theory does not consider the delay characteristic. In thisdissertation, delay constraints are incorporated into the instantaneous transmissionrate based on the concept of effective capacity. The transmission rate based on ef-fective capacity and delay characteristic is formulated as an object function with theconstraint of power control. An optimal rate solution is obtained by applying the con-vex theory. Then a utility function is introduced to combine the instantaneous rate anduser expenditure. Finally the user accesses the network that has the highest valueof the utility function. Both the mathematical analysis and simulations show that theproposed network selection policies can improve network throughput while providingquality-of-service guarantees.3. Combining network conditions with instantaneous transmission rate based on the concept of effective capacity, joint network selection policies are proposed. In thefirst section, a joint delay constraint and resource prediction policy for network se-lection is proposed to manage the problems caused by delay constraint of differentservices and resource uncertainty in integrated wireless systems. The probabilities offree and occupation are evaluated, then the transmission rate based on the conceptof effective capacity is formulated as an object function with the constraint of powercontrol. An optimal rate solution which can provide quality-of-service guarantees isobtained by applying the convex theory. Finally the user accesses the proper networkaccording to both prediction result and transmission rate. Simulation results show thatthe joint network selection policy can provide higher transmission rate than water-fillingpolicy in different fading channels. The transmission rate which is achieved by the jointnetwork selection policy is closer to the upper-bound rate. In the second section, anSINR driven joint network selection policy is proposed to manage the problems causedby delay constraint of different services and intercell interference in integrated wirelesssystems, which incorporate the delay constraint and signal strength into SINR. Thispolicy always allows users to select the network that has a higher SINR value fromall the accessible networks under the delay and signal strength constraint. Both themathematical analysis and simulations show that the joint network selection policy canachieve higher network throughput and SINR with the variety of the delay characteris-tic.4. A new network selection strategy based on auction theory is proposed tomanage the problem caused by the selfishness of the users. In switched networkselection policy, each user selects only a single network. An optimal bid is obtainedaccording to both quality function and interference quality function. Then the networkallocates the resource to the user who gives the highest bid. In paralleled networkselection scheme, each user can access multiple networks, the user who gives thehigher bid will obtain more resources. Both the mathematical analysis and simulationsshow that the network selection strategy based on auction theory can achieve highernetwork throughput and lower outage probability.