Font Size: a A A

Study On The Channel Allocation Method For Mixed Service

Posted on:2009-07-27Degree:MasterType:Thesis
Country:ChinaCandidate:Y S LiFull Text:PDF
GTID:2178360242980360Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
This paper studies the service mainly refers to mixed voice and data services. The two operations have different service characteristics .The characteristic of voice in the actual network is an immediate rejection queuing model, and that of data services is a refused to delay queuing system model.Channel, which is often called the wireless "band", works as a wireless signal transmission medium to transmit the media signal.With the increasing of these two services, the wireless channel which is used to be provided resources has to supply both these two services which are different in service characteristics.How to ensure the quality of service of both these services,wireless channel resources rational distribution has become a task of operators and equipment manufacturers.In this paper,I have done some research in the GPRS/EDGE packet data transmission process and discussed the system of voice and data services channel allocation method.I have proposed a comprehensive cell channel distribution method which distinguishes voice and data services. The aim of the method is to guarantee these two different services under the premise of the quality of service and allocate reasonably of resources. In a word, the main research results are as follows:1. Using the Markov model to establish a channel distribution which is based on the Ericsson equipment.The GPRS/EDGE based on the available GSM network. Data and voice services shares the same limited resources. The voice channel which is vacant can be dynamically allocated to the GPRS/EDGE.This dynamic process can be described by a mobile border Markov process. Based on the Ericsson equipment channel allocation mechanism, I establish a two-dimensional continuous-time Markov chain.Using the method of linear equations, I find a steady-state solutions of Markov chain. Finally by using the numerical calculation, I get the key performance indicators of voice and data services: obstruction of voice, packet losing rate, the average queue length data and data such as the average waiting time targets. By obtaining the numerical solution of key indicators, I do some simulations under the fixed FPDCH Channel allocation and dynamic FPDCH Channel allocation circumstance. The results showed that FPDCH channel allocation should abide by the following principles: the number of FPDCH channel should not increase blindly, it should give full consideration to the impact of the voice service; it should not sacrifice the performance of voice service to access to larger and better utilization of the data service indicators, and it should make full use of the gap voice to transmit data service.The Markov model is established only from the point of channel allocation. It does not distinguish different types of data services and different ways which is used to carry data services. At the same time, the packet losing rate of data services which is based on the Markov model can not accurately reflect the actual network obstruction of data rate. Under the circumstances of actual network which transmit speed is slow, the data users would not wait until their data packets are serviced. They will cut off network connections initiatively. Based on these two point, this paper presents a method which is combined Campbell method and refused to delay the system model, and give a channel allocation method based on the TBF / ETBF queuing data channel allocation method.2.Proposing a model which is a combination of Campbell method and refused to delay the channel allocation methodCampbell method combine with refused to delay the channel model is used Campbell method to transform of the service of each user data throughput equivalent to a virtual service Erl volume. Considering service level GoS voice and data services QoS service quality static and dynamic sub-channel distribution. In this method, in accordance with the Campbell Little theorem and multi-service access to the equivalent of thinking and method of synchronization Campbell refused to use the delay-line model.Campbell method takes into account a variety of data comprehensively which is from the point of view of user-aware. It is a good channel allocation method. However, it is strict in data collection. Mainly in the core network service side of distinction, but in the core network side which is to distinguish between services will not be able to distinguish between loads. The load distinction manner could not differentiate services. In order to distinguish between both service and Carrying way to differentiate wireless side of the statistical data conversion. Campbell method is strict in the data collection. In the actual network data acquisition is difficult, the paper do some study mainly from the calculation of the theoretical methods.3.Proposed a data channel allocation method which is based on the TBF/ETBE queuing model firstly.Data channel allocation method which is based on the TBF/ETBE queuing model gives full consideration to the feasibility of collecting data. From the GPRS/EDGE network daily maintenance statements , we can establish a complete set of channel allocation method.The method is composed of three main parts: voice channel configuration, data channel allocation and static and dynamic channel allocation. (1) Voice channel allocation is the adoption of the classic Erlang B formula. Traffic will be converted back to voice channel number. (2) Data channel configuration is considering GPRS/EDGE wireless side can not statistical data users. Taking TBF/ETBF as an "actual user" to transmit data. With the wireless side of the statistical data related to establish a complete set of TBF/ETBF queuing theory. In the theoretical system, some of the service as channel can be seems as a service window of the TBF/ETBF services. By using refused to delay adoption of the M/M/m(n) queuing formula, I access to TBF/ETBF services window of the service. Then converted back to the number of service channel. (3) The static and dynamic channel allocation gives priority to the use of EDGE data services strategy. Finally I get the number of the TCH Channel and that of the FPDCH Channel .The method adopted two means of verification, one is statistical data validation .the other is verification area measured. The validation results show that the method considered the voice and data services of the respective provisions of the quality of service, distribute the resources reasonably. Mixed service-oriented channel allocation method has made some progress in recent years. But due to its relatively short history, theory and methods are constantly improving. In this paper, I do lot of research on the GPRS/EDGE System voice and data services channel allocation methods. Finally found that the channel allocation which is based on the TBF/ETBF queuing data channel allocation method can be easily applied to the actual wireless network planning and optimization. The method also gives the relevant theoretical basis for the study of wireless channel allocation method for the forthcoming 3G commercial communication systems.
Keywords/Search Tags:channel allocation, mixed service, Markov model, GPRS/EDGE, Campbell, TBF/ETBF
PDF Full Text Request
Related items