Scheduling algorithm for multi-item requests with time constraints in mobile computing environments

Abstract

On-demand broadcast is an effective wireless data dissemination technique to enhance system scalability and capability to handle dynamic user access patterns. Previous studies on time-critical on-demand data broadcast were under the assumption that each client requests only one data item at a time. Little work, however, has considered the ondemand broadcast with time-critical multi-item requests. In this paper, we study the problem arising in this new environment and observe that existing single item based scheduling algorithms are unable to manage multi-item requests efficiently. Thus, a new scheduling algorithm that combines the benefit of data item scheduling and request scheduling is proposed. The performance results show that the proposed algorithm is superior to other classical algorithms under a variety of factors. Our algorithm not only reduces deadline-missing ratio of requests, but also saves broadcast channel bandwidth.

On a novel property of the earliest deadline first algorithm

“…Realtime data broadcast, an indispensable component of mobile real-time data services, addresses data dissemination that arises when there are timing constraints imposed on data such as stock quotes, traffic information and dynamic news webpages, etc., or when there are queries issued by "impatient" users [10]. Existing real-time data broadcast research mainly focuses on how to generate broadcast scheduled to meet timing constraints of single data requests [11], [12] or one-shot queries [13], [14]. In an important class of data broadcast applications, however, mobile users are interested in monitoring multiple time-varying data items continuously in order to be kept informed of up-to-date information.…”

A Novel Scheduling Algorithm for Supporting Periodic Queries in Broadcast Environments

“…Finally, we order the newly generated task set T P List in non-decreasing order of pp (line 13) and return the result. The time complexity of Algorithm 5: Admission Control of UM input : OQList output: Scheduling sequence BS of queries in T P List 1 begin 2 Generate 2-harmonic task list T P List of OQList by Sr and compute anum for each T P ; 3 Invoke Algorithm MQM (with T P List as input); 4 Invoke Algorithm RQM (with T P List derived in the last step as input); 5 if utilization of T P List ≤ 100% then 6 Use RM to generate the scheduling sequence BS of T P List within one hyper-period; Based on Definition 4, we know τ i is an R-task (line 4). Next, we traverse T P List with cursor j to find the RM -task τ j of τ i , and merge them (lines 5 − 9).…”

“…Based on the definition of R-task and the rules of item division, we know T j > T i . Because τ i and τ i+1 are Algorithm 2: RQM-UO input : a 2-harmonic task set, the current time t output: the data item to be broadcast at t 1 begin 2 Utilize the RM algorithm to derive the task schedule for the 2-harmonic task set within one hyper period; 3 Find the task to be executed at time t; 4 if the task to be executed is a virtual task then 5 Find the R-task and RM -task of the virtual task; 6 if broadcasting the data item accessed by the R-task is redundant then 7 Return the data item accessed by the RM -task; 8 else 9 Return the data item accessed be the R-task; 10 else 11 Return the data item accessed by the task; two adjacent DC tasks, and the DC-tasks are sorted in nondecreasing order of period, we know j must be larger than i, which indicates that T j ≥ T i+1 . Since τ i is an R-task, based on Definition 4, we have,…”

“…A time-critical request is associated with a deadline imposed by either the application or the user, the result of a request is useful only if all the required data items can be received before the deadline. For such kind of applications, some real-time broadcast scheduling algorithms have been proposed [14], [6], [22], which take queries' deadlines into account. But unfortunately, almost all the existing real-time broadcast scheduling algorithms only support single data item [2] or one-shot queries [7], [16], and thus can be of low performance, or even not applicable in some practical applications.…”

Exploiting replication on dependent data allocation for ordered queries over multiple broadcast channels

“…Moreover, the proposed scheduling algorithm is more scalable than LTSF, and hence, is suitable for practical use. In recent years, several researches are focused on scheduling for dependent data in on-demand broadcast environments [12,14,17,44]. In addition to devising scheduling algorithms, several studies consider the employment of data indexing in on-demand data broadcasting environments [21,24,33].…”