Chapter 1 Introduction

Transcription

1 Chapter 1 Introduction 1.1Motivation The past five decades have seen surprising progress in computing and communication technologies that were stimulated by the presence of cheaper, faster, more reliable electronic components in the market. The design of smaller and more powerful devices enabled their mobility, which is quickly changing the way we compute and communicate. Wireless and mobile networks are emerging as networks of choice, due to the flexibility and freedom they offer. The use of satellite, cellular, radio, sensor, and ad hoc wireless networks, wireless local area networks (LAN), small portable computers, and personal communication systems (PCS) is increasing. These networks and devices support on the move computing trend, known as mobile computing, nomadic computing, or computing anywhere anytime. The applications of mobile computing and wireless networks include e-commerce, personal communications, telecommunications, monitoring remote or dangerous environments, national defense (monitoring troop movements), emergency and disaster operations, remote operations of appliances, and wireless Internet access. Explosive growth has been witnessed during the last few years in the demand for mobile communication services, particularly in cities [146]. From available studies, it is clear that this rate of increase in mobile communication services will continue for quite some time [7]. In order to make an arrangement for this projected increase and solving the present crisis of channel scarcity for mobile radio use, some improvements to mobile wireless systems have been recommended. 1.2Cellular Architecture One of the favored and most frequently used techniques for increasing the capacity or efficiency of frequency spectrum utilization is the implementation of a cellular architecture in mobile communication. In the cellular architecture approach, the entire geographical area is divided into cells or zones and each cell is serviced by a mobile 1

2 service station (MSS), which is located at the center of each cell. Fig. 1.1 shows a 6 x 6 grid cellular system. In the cellular systems, instead of using one powerful transmitter, many low power transmitters are deployed throughout the coverage area. Each transmitter talks to many mobiles at once, using one channel per mobile. Channels use pair of frequencies for communication, one frequency for transmitting the call and another frequency for receiving the call. When a mobile device, also called a mobile host (MH), wants to start a communication session with another MH, it sends a connection request to the nearest MSS through a control channel. After receiving the request, the MSS searches for a free radio channel which must not be in use in that cell or in neighboring cells, otherwise interference of signals would occur. If such a free radio channel is found, the MSS will inform the MH to use it. Then, the MH starts sending and receiving data packets through the selected free channel, and the MSS will forward those packets to and from other parts of the wired network Reuse Distance The same channel can be concurrently used by other transmitters if they are separated by a distance called minimum channel reuse distance. In order to give a wide coverage to the mobile users, the cellular network consists of thousands of cells and the network is further split into clusters and the frequency allocations done in one cluster can be applied to other clusters in that network. Fig 1.1: 6 x 6 Grid 2

3 1.3Channel Allocation Increase in demand and the poor quality of existing service led mobile service providers to research ways to improve the quality of service and to support more users in their systems. Because the amount of frequency spectrum available for mobile cellular use was limited, efficient use of the required frequencies was needed for mobile cellular coverage. This requires an efficient channel allocation scheme. The task of a channel allocation scheme is to allocate radio channels to the cells or mobiles in such a way as to minimize call blockings or call droppings, and also to maximize the quality of service (QoS). The channel allocation schemes can be classified into three categories: fixed channel allocation (FCA), dynamic channel allocation (DCA), and hybrid channel allocation. In fixed channel assignment (FCA), radio channels are permanently allocated to each cell for its exclusive use and allocations is done according to traffic load estimation, cochannel and adjacent channel interference constraints. The traffic information is very difficult to predict due to the mobility of users. Therefore, FCA scheme is not frequency efficient because channel allocation cannot adapt according to the dynamically changing distribution of mobile terminals in the cell or area. In order to overcome the problems associated with FCA, various traffic adaptive channel allocation schemes have been developed such as dynamic channel assignment (DCA) and hybrid channel assignment (HCA). The DCA scheme is of two types: centralized and distributed. In centralized DCA, all channels are kept in a pool managed by a central controller. The related base station will ask the central controller for a channel when a new call request arrives and when the call is completed, the channel is returned to the central channel pool. In distributed DCA, a radio channel is chosen by the local base station of the cell where the call is initiated. A channel is suitable for use in any cell provided that co-channel and adjacent channel signal interference constraints are satisfied. As more than one channel 3

4 may possibly be available in the channel pool to be allocated to a call when required, some approach must be applied to choose the allocated channel. Although the DCA schemes can adapt channel allocation to dynamic traffic loads, it can also considerably increase the network complexity due to co-channel cell locking and other channel management because it is a call-by-call based allocation. In order to maintain both cochannel and adjacent channel interference below a certain threshold, cells within the required minimum channel reuse distance from a cell that borrows a channel from the central pool cannot use the same channel. DCA also requires fast real-time signal processing and associated channel database updating. A compromise between the radio spectrum efficiency and channel management complexity is HCA, which combines FCA with DCA. In HCA, all available channels are divided into two groups, FCA group and DCA group, with an optimal ratio. It has been shown that both DCA and HCA can achieve a better utilization of radio channel resources than FCA in a light traffic load situation, due to the fact that both schemes can adapt to traffic load dynamics. However, they may perform less satisfactorily than FCA in a heavy traffic load situation due to the necessary channel locking Channel Borrowing Another approach to adaptive channel assignment is channel borrowing, in which the channel resources are divided into borrowable and non-borrowable channel groups. The non-borrowable group is assigned to a cell in the same way as FCA, When all of its fixed channels are occupied, a cell borrows channels from its neighbor cells which have a light traffic load. More recently, a channel borrowing scheme called channel borrowing without locking (CBWL) is proposed, where the C channels of each base station are divided into seven distinct groups. The C 0 channels of group 0 are reserved for exclusive use of the given cell. The (C i, i = 1, 2,., 6) channels of the other six groups can be borrowed by the six adjacent cells respectively, one group by one adjacent cell. Each borrowing channel is used with a limited power level. That is, the borrowed channel is directionally limited as well as power limited. Therefore, the channel locking for cochannel cells is not necessary. 4

5 1.4Problem Specification Even though, the growth of mobile communication users is very rapid, the radio channel available for communication is very limited. Thus, for supporting large number of mobile users, efficient use of limited radio spectrum is required. In order to reuse radio spectrum, wireless systems use cellular architecture for efficient use of spectrum. A radio channel i can be used by a cell c without any interference, if it is not simultaneously used by any other cell within the minimum reuse distance of cell c. Thus, the problem of channel allocation is to design architecture and a scheme for allocating channels to cells in order to eliminate co-channel, adjacent channel, co-site interference, for mobile calls. Considerable efforts have been devoted to provide interference free calls, but some issues have been neglected because of complexity involved in designing optimum algorithms. 1.5Objectives The objective of this research is to minimize the call blockings and call droppings with the efficient use of spectrum, and this can be made possible with the design and implementation of efficient hybrid channel allocation algorithms for channel allocation. To achieve this, a new strategy in channel allocation schemes is proposed which uses Agent and Multi-Agent approach to make dynamic decision and do the computation in the remote destination, in order to reduce the network traffic and improve the efficiency of resource allocation. The contributions of this research are: Minimize blocked calls Minimize dropped calls Minimize channel acquisition time. Maximize the number of communication sessions Ability to adapt to changing load The effort to adapt to the above requirements resulted in the development of Optimized blocking dropping load balancing (OBDLB) algorithm. 5

6 1.6Organization of the Thesis The remainder of the thesis is organized as follows. Chapter 2 provides an introduction of cellular concept. Various ways of frequency planning, frequency reuse and radio channel allocation are also shown. It also explains various variables that affect the capacity of a cellular system. Chapter 3 presents a comprehensive survey of the channel allocation problem. The large number of existing channel allocation methods are classified, which categorize the various channel allocation methods based upon the techniques used by the system to obtain network information. Different agent architectures are also discussed in this chapter. Chapter 4 explains the Hybrid Multi-Agent Architecture that is applied to solve the channel allocation problem. Functionality of each layer of INTERRAP agent architecture is also provided, along with the Optimized Blocking Dropping Load Balancing (OBDLB) Algorithm to find a solution of channel allocation problem under different traffic loads. Chapter 5 includes the results of OBDLB algorithm, which are implemented on multiagent systems. The proposed scheme is also compared with the existing schemes i.e. Fixed Channel Allocation and Dynamic Channel Allocation. Chapter 6 draws some conclusions suggested by the experimental analysis and point out some possible future developments in the research in this field. 6