Research On Channel Allocation Scheme For Heterogeneous Wireless Networks

With the rapid development of modern technologies, mobile communication exertsmore and more important effect. Nowadays the commercial technology of3G is indeployment, and the research for next generation mobile communication, i.e.,Beyond-3G (B3G) is in the ascendant. Instead of using a single technology as a symbol,B3G will be an omnidirectional evolution of access technology, network structure andsystem performance, which is an all-IP heterogeneous telecommunication network. Theheterogeneity brings various fundamental challenges. In order to offer ubiquitouswireless access at any time, the interworking and harmonization among heterogeneousnetworks are considered as the primary problem in the coming years. Thus the RRM(Radio Resource Management) of all-IP heterogeneous network has got lots of attentionand becomes a valuable research task.RRM is employed to utilize the radio resource reasonably and effectively. Itinvolves in several aspects, e.g. power control, mobility management, channel allocation,admission control, etc. In the view of simulation models, schemes and algorithms, theresearch of channel allocation for all-IP heterogeneous network is performed in thisdissertation. By combineing several channel allocation strategies or methods, somechannel allocation schemes are proposed for wireless networks, with the objective tolower the service blocking probabilities and increase the radio resource utilization ofwireless systems. Based on analytical model and computer simulation, the proposedschemes and algorithms are compared with existing schemes, which maybe of somevalue for the future research of RRM.The contributions are listed as follows.Firstly, the traffic model for wireless multimedia networks is studied in thedissertation. Considering the characteristics of person’s mobility and the subscriber’sdensity change on different time in a day, channel allocation schemes combined withtwo pre-emptive priority strategies (PD scheme and PV scheme) are proposed. Theproposed schemes are discriminated by the different kinds of pre-empted services. Byusing two special periods as the example, the traffic characteristic of time is alsoanalyzed and considered in the traffic model. For analytical simplicity, an approximatemodel is introduced. From the simulation results and analysis, some conclusions can beobtained that PV strategy outperforms PD strategy if the video services can betransformed into non-real-time services and this conversion can be accepted by users.Secondly, in this dissertation channel borrowing based channel allocation schemefor heterogeneous wireless networks is proposed. By utilizing the queue theory forclosed queuing networks, the scheme is analyzed from the viewpoint of connection-level QoS. The contributions of this part lie in three aspects. First, based on the horizontalchannel borrowing, the vertical channel borrowing is introduced in this channelallocation scheme, and the channel borrowing is restricted within handoff requests, andonly pre-defined channels can be lent. When all the channels in a target cell are occupied,handoff requests can borrow channels from the adjacent homogeneous cells. In case thatthe horizontal borrowing requests fail, handoff requests may also borrow channels fromheterogeneous cells. Second, to further improve system performance, otherpriority-based strategies such as channel reservation and pre-emptive priority areintroduced. And the third, to meet the high QoS requirements of some services andlower the blocking probabilities, certain type of services can be transformed betweenreal-time services and non-real-time services as necessary. The proposed scheme is alsocompared with CBWL scheme. The simulation results demonstrate that the proposedschemes can lower the service blocking probabilities and improve channel utilization.Thirdly, the channel allocation scheme based on channel borrowing and trafficoverflowing for heterogeneous wireless networks is studied in this dissertation. Bysynthetically considering both horizontal channel borrowing and vertical trafficoverflowing, a new channel allocation scheme (HCB-VTO) is proposed. To lower theservice dropping probability, handoff requests are allowed to borrow channels fromhomogeneous cells or overflow to heterogeneous cells in the case there are no idlechannels in the target cell. Considering the different priorities among services, somekind of services with high priority are allowed to pre-empt the channels originallyemployed by other services. To further decrease the service dropping probabilities,waiting queues are introduced and the handoff requests of different services can bequeued. Two cases are analyzed in this part and the corresponding two HCB-VTOschemes are presented: the heterogeneous network composed of macro-cells andmicro-cells, and the heterogeneous network composed of macro-cells and hotspots. Thesimulation results prove that the proposed schemes can lower blocking probabilities ofboth the originating service and handoff services, and then system performance isimproved.Finally, policy-based radio resource allocation scheme for wireless multimedianetworks is studied in this dissertation, in which the characteristic of wireless traffic isanalyzed and the corresponding traffic model is described. To improve systemperformance, different channel allocation algorithms and channel allocation mechanismsare introduced, and then a series of policies are formed by combining them. Thesepolicies are put into a policy list and a policy-based channel allocation scheme isconstituted. The core of policy-based channel allocation scheme lies in two aspects: thechannel allocation mechanism (for example DCA or FCA) is selected according tonetwork load; different channel allocation algorithms (for example PGC or PSCR) arechosen in terms of pre-defined scenarios. The aim to improve the radio resourceutilization and to lower the blocking probabilities is achieved by allocating wirelesschannels flexibly and dynamically. The simulation results are also demonstrated that the introduction of policy-based method can improve the system performance comparedwith the single channel allocation strategy.