Optimal system capacity in handover prioritised schemes in cellular mobile telecommunication systems

Abstract

This paper investigates the issue of capacity maximisation in the presence of new call and handover requests in cellular mobile telecommunication systems. Initially, both non-prioritised and prioritised handover schemes are analysed leading to the identification of the optimal configuration (e.g. number of reserved channels, queue length) in order to maximise the system capacity. The ‘guard channels’ prioritised scheme is analysed in the sequel and it is proved that it does not maximise the system capacity, however, it approximates the optimal configuration. Finally, a new handover prioritised scheme, which achieves the maximum system capacity, is proposed and analysed. This scheme is based on a new resource management scheme: the ‘unequally shared channels’, which are allocated to new calls and handovers with a different priority determined by a single control parameter.

Introduction

The capacity of a cellular system, apart from the offered traffic load, is influenced by the user mobility. This is because each base station should serve, apart from new call requests, handover requests as well. Certain quality of service requirements impose the need to guarantee a lower blocking probability for handover requests compared to the relevant new call blocking probability. In addition to this requirement, and taking into account that a call may face a series of handovers throughout its duration, it is also necessary to guarantee that the probability of an ongoing call being dropped during a handover attempt is kept under tolerable levels. Such requirements in fact limit the system capacity in terms of the maximum offered load that the network can handle without violating the quality of service constraints.

In second generation systems, where the number of users is relatively low and the cell size relatively high, the effect of mobility on the system capacity is rather weak and it is not actually taken into account during the system dimensioning. In third generation systems, however, the number of subscribers will be enormous and the cell size will decrease making thus the impact of mobility much stronger. Micro-cells have a radius of a few hundreds of meters while even the macro-cells in densely populated areas are expected to have a radius of less than one kilometre. In such an environment the amount of handovers that the system has to handle will be significantly high, while the increased number of handovers per call indicates the need for incorporating the influence of mobility in the air-interface dimensioning process.

The issue of traffic modelling in cellular mobile systems, including the aspect of mobility, has been extensively analysed in the literature [1], [2], [3], [4], [5], [6], [7], [8]. The key requirement, that the handover blocking probability should be lower than the new call blocking probability, has led to the development of handover prioritised schemes which aim at the maximisation of the system capacity under certain quality of service constraints regarding both handover and new call blocking. Handover prioritised schemes utilise two main techniques for providing priority to handovers: (a) handover reserved channels; and (b) queuing handover priority. In the first case, a set of channels are exclusively utilised for serving handover requests while the rest of the channels are available for both handovers and new calls. In the literature [1], [2], [3], [4], [5], [6], [7], [8] this scheme is referred to as the cutoff priority scheme (CPS) or the guard channel scheme. In the second case, the presence of a queue allows for some prioritised management scheme of the queued handover requests. Although handoff-prioritising schemes provide improved performance, they do that at the expense of a reduction in the total admitted traffic and an increase in the blocking probability of new calls. In order to cope with that, other prioritised schemes allow for the new calls to be queued too. Of course, a combination of both techniques (i.e. reserved channels and queues) is also valid. The proposed handover prioritised schemes succeed in increasing the system capacity while under a certain configuration of the scheme control parameters (e.g. number of reserved channels, queue length, queue time out, etc.) it is possible, in some cases, to achieve a maximum system capacity.

In this paper, we analyse the issue of the system capacity initially assuming a generic handover prioritised scheme. The analysis leads to a certain condition under which the system capacity is maximised. As proved in the sequel the generic handover prioritised scheme for which the optimal capacity condition may not be feasible to be fulfilled, there is a certain methodology for identifying the highest possible system capacity that the generic handover prioritised scheme can achieve. This methodology proves to be a generalisation of the algorithm followed in Ref. [1]. Based on these results, the handover-prioritised schemes proposed in the literature are analysed in terms of their capability of achieving the maximum system capacity. As it is proved, depending on the nature of the prioritised scheme, the condition of maximal system capacity is not actually achieved, leading thus to an approximation of the maximum capacity.

As a final step of the analysis a new handover prioritised scheme is proposed based on a new concept: the unequally shared channels. According to the proposed scheme, there is a set of channels that are available to both handovers and new calls, however, not with equal priorities. The proposed scheme is quite simple as a unique control parameter determines the overall operation. As it is proved the proposed scheme under an appropriate configuration achieves the condition of maximum system capacity. Analytical results based on a Markovian modelling approach provide an acknowledgement of this conclusion.

The overall analysis, apart from the definition of an efficient handover prioritised scheme which maximises the system capacity, provides the means to analyse the efficiency of the various proposed handover prioritised schemes and to provide their optimum configuration so as to approximate or achieve the maximum system capacity.

It should be noted here that the analysis presented in this paper adopts a set of assumptions (e.g. a fixed number of channels per cell, one channel per call, etc.) which are valid for second generation systems, however, they might not be the case in third generation systems where the operators may decide to apply some dynamic channel allocation scheme. In this context, the analysis presented in the current paper can be considered as an initial step towards the more general problem of the system capacity optimisation in dynamic channel allocation schemes.

This paper is organised as follows: Section 2 deals with the air-interface dimensioning in cellular mobile systems and provides the QoS constraints that affect the system capacity. Section 3 analyses the non-prioritised radio resource management schemes in terms of capacity. Section 4 provides a theoretical framework regarding the optimal capacity in handover prioritised schemes. Section 5 considers prioritised handover schemes with reserved channel while in Section 6 a handover prioritised scheme which is proved that it can achieve the maximum system capacity under appropriate configuration is proposed. Section 7 summarises the conclusions of the paper.