Improved minimum latency aggregation scheduling in wireless sensor networks under the SINR model

Abstract

Wireless sensor networks are attracting much attention from the world and Minimum Latency Aggregation Scheduling(MLAS) has become one of the most significant fundamental problems in wireless sensor networks. However there are few results about efficient data aggregation algorithms under the Signal-to-Interference-plus-Noise-Ratio (SINR) model. In this paper, we propose a centralized algorithm to aggregate data from all sources in O(log 2 n) time slots where n is the total number of nodes. To the best of our knowledge, this is the current best result for the problem. This algorithm uses round scheduling, topology construction and non-linear power assignment as the main techniques. We give a detailed proof of correctness, also an aggregation latency analysis of the algorithm as well as the parameter constraints to achieve our result.

Minimum Latency Link Scheduling for Arbitrary Directed Acyclic Networks Under Precedence and Sinr Constraints

“…An example of precedence is when a node needs to compute the maximum or average value of all the collected data from other nodes. Very recently, there has been some work on this kind of minimum latency aggregation scheduling problem (MLAS) [7,16,21,26,29,30,32,40,43,44], but (1) except the work in [16,21,29,30,32], most of them study the scheduling problem based on the graph based interference models instead of the SINR model; and (2) all of the current work study the joint problem of tree topology construction and link scheduling where each sensor only needs to send their aggregated data to one parent node. In this paper, we study the problem of minimum latency link scheduling for arbitrary directed acyclic networks under both precedence and SINR constraints (MLSDAN).…”

“…Compared with the MLAS problem, the MLSDAN problem allows multiple sinks whereby each sensor may need to transmit data to more than one parent node. In addition, in the MLAS problem, given the nodes arbitrarily located on the plane, the algorithm can take advantage of the specifically constructed tree topology to facilitate the scheduling process (e.g., [16,21,29,30,32]). Whereas in our MLSDAN problem, an arbitrary network topology is given and thus the algorithm loses the freedom of constructing its own network topology to help the link scheduling process.…”

“…Xu et al have also studied the MLAS problem in the protocol interference model [43]. Currently, the only results for the MLAS problem under the SINR model can be found in [16,21,29,30,32]. All of these work involve their own tree topology construction algorithms in order to expedite the link scheduling process.…”

“…Basically, this paper shows that a connected wireless topology can be constructed in O(log 4 n) timeslots where n is the number of nodes. This paper has spawned a large body of following works that either aim to reduce the scheduling lengths [4,13,16,21,23,[34][35][36] or to analyze the hardness of the scheduling problem [12,15,27]. In addition, since the non-linear power assignment based scheduling algorithm needs the information of all the links such as their lengths, there are some recent works that focus on oblivious power assignment methods.…”

“…Precedence constraints are common in data aggregation problems where a sensor can not send data to its parent node before it has received data from all of its children. Existing solutions to the so-called minimum latency aggregation scheduling problem [7,16,21,26,29,30,32,40,43,44] mainly focus on specific tree topologies rooted at the sink node. In this paper, we study the minimum latency link scheduling problem for arbitrary directed acyclic networks under both precedence and SINR constraints.…”