在环境恶劣且无固定Sink的无线传感器网络,节点为了避免意外死亡而丢失数据,需要能量有效地将数据分发到其他一部分节点上存储,以等待移动Sink进行采集.提出了一种新的基于Luby变换码(Lubytransformcodes)、简称LT码的数据持续性提高方案(LT-codes based scheme for improving data persistence, LTSIDP),其中LT码是一类纠删码.LTSIDP将数据的存储过程分为2个步骤:第1步,节点根据一段时间内接收数据包的情况估计网络中数据包的数量和节点的总数,进而计算出基于LT码存储数据所需要的参数;第2步,节点再根据获得的参数对接收到的数据进行存储.每一轮LTSIDP执行结束后,移动Sink可以在一定时间段内的任意时刻和任意地点进入网络,访问少量仍然存活的节点就能获得所有源数据.理论分析和实验表明,LTSIDP不仅能获得比已有算法更高的数据持续性,而且能量更有效.
For a wireless sensor network that does not has a fixed sink and is deployed in Hash environment, each node should disseminate its data to a subset of nodes in the network for storage. By this way, a mobile sink can collect all data even if some nodes die due to accident. A novel LT (Luby transform)-codes-based scheme, named LTSIDP, is proposed, where LT codes are a kind of erasure codes. In LTSIDP, the process of data storage is divided into two steps. In the first step, nodes estimate the number of nodes in the network and the number of data packets by receiving packets for some time. After they acquire the numbers, they can compute a parameter of LT codes. In the second step, nodes store the data they received according to the parameter. After LTSIDP terminates at each round, when a mobile sink enters the network at anytime and anywhere in a given interval, it can collect all data even if it just visits a small number of alive nodes. Theoretical analysis and experiments show that LTSIDP can not only achieve higher data persistence but also has more energy efficiency than previous schemes.