Research On Data Forwarding Scheduling For Low-Duty-Cycle Sensor Networks

Different from the traditional wireless ad-hoc networks, energy efficiency opti-mization is usually the main design objective for wireless sensor networks. In order to reduce the energy waste caused by idle listening, wireless sensor network are usu-ally operated with low-duty-cycle mode. However, the low-duty-cycle operation for each node makes the low delay and energy efficient data forwarding be a challeng-ing issue. The low-duty-cycle operation improves the energy efficiency but yields a notable sleep latency. In addition, the low-duty-cycle operation makes wireless media lose the inherent broadcasting nature and thus leads to a low efficiency for broadcasting. Therefore, the data forwarding scheduling problem for low-duty-cycle sensor networks is an important and challenging research issue.The data forwarding scheduling problem for low-duty-cycle sensor network-s mainly focuses on how to design an efficient data forwarding strategy based on known information at link layer, or how to combine with the design of link layer, for low-duty-cycle sensor networks, so as to overcome the challenges of delay-energy tradeoff and energy efficiency for broadcasting, which are derived from the low-duty-cycle operation. Lots of existing works about this problem are proposed, however, they still have many shortages. For example, very few of them considers the trade-off optimization between data latency and network energy efficiency; they do not utilize the broadcasting spatiotemporal characteristic of wireless media to improve the energy efficiency for broadcasting and lack provable performance guarantee for their proposed solutions; they overlook the energy fairness problem of broadcasting and the works about the distributed solutions for broadcast scheduling are lacked. Based on these shortages, this dissertation proposes several efficient data forwarding scheduling schemes. The main contributions of this dissertation are summarized as follows:1. Aiming at the moderate-scale sensor networks in which the events are rare but urgent, a fully distributed Low-delay Energy-efficient Decision Strategy (LED-S) is proposed for selection of the event-reporting candidate. Once an event occurs, the decision-maker will first be found in a distributed way, and then the decision-maker will determine the event-reporting candidate by making an on-line cooperation with its neighboring sensing nodes, which avoids the redundant transmissions and achieves a good network performance involving delay and energy efficiency. For the large-scale networks, moreover, a solution that combines LEDS with a sink augmentation scheme is proposed. Finally, the simulation results demonstrate that LEDS exhibits a satisfactory tradeoff between delay and energy efficiency. and also outperforms the existing solu-tions.2. Considering how to utilize broadcasting spatiotemporal locality to address the broadcast scheduling problem in low-duty-cycle WSNs. First, the target prob-lem is transformed into the Latency-optimal Group Steiner Tree Problem on the Spatiotemporal Relationship Graph, which is shown to be NP-hard, and then this problem is approximately solved by using a deterministic randomized-rounding based method. Also, a novel Broadcasting Schedule Construction Algorithm is proposed to further avoid the redundant transmissions and re-duce the collision probability as much as possible. Finally, the high-efficiency of the proposed solution is evaluated through both theoretical analysis and simulations.3. Considering the Load-Balanced Minimum Delay Broadcast Scheduling Problem (LB-MDBS) for low-duty-cycle WSNs. First, LB-MDBS problem is trans-formed into the equivalent Load-Balanced Parents Assignment Problem (LB-PA), and its NP-hardness is proved. Then, this problem is addressed by propos-ing the Load-Balanced Parents Assignment Algorithm (LBPA-A) which can provide a provable approximation ratio. Also, an efficient distributed solution is presented. Finally, the high-efficiency of the proposed solutions has been evaluated by simulations in terms of energy fairness, total energy consumption and delivery ratio.Finally, the dissertation is concluded with possible future research works dis-cussed.