A complementary approach to data broadcasting in mobile information systems

Abstract

Acharya et al. have proposed the use of a periodic dissemination architecture in the context of mobile systems, called Broadcast disks. This strategy can construct a memory hierarchy where the highest level contains a few items and broadcasts them with high frequency while subsequent levels contain more and more items and broadcast them with less and less frequency. In this way, one can establish a trade-off between access time for high-priority data and that of the low-priority items, where access time means the time elapsed from the moment a client submits a query to the receipt of data of his (her) interest on the broadcast channel. However, based on Acharya et al.'s algorithm, some broadcast slots may be unused, which results in the waste of bandwidth and the increase of access time. Therefore, in this paper, we propose an efficient broadcast program, the complementary approach, in which no empty slot is wasted. The basic idea of the complementary approach is to move some pages which are located near the end of a broadcast cycle to those empty slots which occur before these pages. Therefore, finally, the total number of slots in a broadcast cycle is equal to the one computed from Acharya et al.'s algorithm minus the number of empty slots. Obviously, our complementary approach generates a small number of slots in one broadcast cycle and shorter mean access time than Acharya et al.'s algorithm.

Introduction

The emergence of powerful portable computers, along with advances in wireless communication technologies, has made mobile computing a reality [2], [9]. In the evolving field of mobile computing, there is a growing concern to provide mobile users with timely access to large amount of information [8], [11], [19]. Examples of such services include weather, highway conditions, traffic directions, news and stock quotes.

Although a wireless network with mobile clients is essentially a distributed system, there are some characteristic features that make the system unique and a fertile area of research [2]. These features include: asymmetry in the communications, frequent disconnections, power limitations and screen size [2]. Among them, asymmetry in the communications means that the bandwidth in the downstream direction (servers-to-clients) is much greater than that in the upstream direction. The communication asymmetry, along with the restriction in power that the mobile units have, has made the model of broadcasting data to the clients, an attractive proposition.

In traditional client–server information systems, clients initiate data transfers by sending requests to a server [1], [10]. Such systems are pull-based, which has the advantage of allowing clients to play a more active role in obtaining the data they need. However, pull-based systems are a poor match for asymmetric communications environments, as they require substantial upstream communications capabilities. Therefore, another push-based architecture that exploits the relative abundance of downstream communication capacity in asymmetric environments is proposed [13].

In a push-based information system, servers broadcast the desired data items in the broadcast channel continuously and repeatedly. The main advantage of broadcast delivery is its scalability: it is independent of the number of users the system is serving. In the simplest scenario, given an indication of the data items that are desired by each client listening to the broadcast, the server would simply take the union of the requests and broadcast the resulting set of data items cyclicly, as was done in Datacycle [5], [7]. Therefore, retrieving data pages from the broadcast channel can be viewed as sequentially accessing the broadcast data, where access time is the amount of time a client has to wait for an information item that it needs. It is important to minimize the access time so as to decrease the idle time at the client [6]. Alternatively, the server can broadcast different items with differing frequency [1], [10]. Such a broadcast program can emphasize the most popular items and de-emphasize the less popular ones, which was proposed in Acharya et al.'s broadcast disks [1], [10].

There have been many strategies proposed for an efficient broadcast delivery. Basically, these strategies can be classified into two types: static and dynamic. By “static broadcast”, we mean that a broadcast where the schedule of programs is fixed and even though the contents of a program can change with time [1], [10]. In contrast, in “dynamic broadcast”, both the schedule of programs and its contents can change and there exists limited support to handle user's requests [12], [14], [15], [19]. Also, there have been some researches on reducing access time [1], [10], nonuniform broadcast [1], [10], [16], [17], fault-tolerance [3], [4], and broadcast data on multiple wireless channels [18].

Among those strategies for efficient broadcast delivery, Acharya et al.'s broadcast disk strategy [1], [10] is one of the well-known static algorithms. Based on broadcast disks, the server can construct a memory hierarchy in which the highest level contains a few items and broadcasts them with high frequency while subsequent levels contain more and more items and broadcast them with less and less frequency. However, based on Acharya et al.'s approach, some broadcast slots may be unused, which results in the waste of bandwidth and the increase of access time. Therefore, in this paper, we propose an efficient broadcast program, the complementary approach, in which no empty slots are wasted. The basic idea of the complementary approach is to move some pages which are located near the end of a broadcast cycle to those empty slots which occur before these pages. Therefore, finally, the total number of slots in a major cycle is equal to the one computed from Acharya et al.'s algorithm minus the number of empty slots. Obviously, our complementary approach generates a small number of slots in one broadcast cycle and shorter mean access time than Acharya et al.'s algorithm.

The rest of the paper is organized as follows. In Section 2, we give a brief description of Acharya et al.'s algorithm and show an example of the empty slot problem in Acharya et al.'s algorithm. In Section 3, we present our complementary approach to solve the empty slot problem. In Section 4, we study the performance of our complementary approach, and make a comparison with Acharya et al.'s algorithm. Finally, Section 5 gives the conclusion.