مقاله انگلیسی رایگان در مورد بازیابی داده های پراکنده با استفاده از بهینه سازی – الزویر ۲۰۱۷

Compressed sensing (CS) framework is aimed at precisely obtaining a high dimensional signal x such that x ∈ Rm×n from a relatively small sample set of linearly combined observations y such that y ∈ Rm, given the linear measurements are the projections of the original signal which is sparse in some intitutively known domain. CS based compression is driven by the idea of obtaining sparse signals at a rate notably lesser than the Nyquist rate. The CS principle utilizes the sparse nature of the wireless sensor network (WSN) data for efficient compression and recovery, However, in practical scenarios the signal in question is not sparse itself but in most of the cases has a sparse representation in some a × b directory ‘D’ such that (a < b). Given the inherent energy and bandwidth constraints of WSNs, CS based data gathering has gained the desired attention in recent years. The central idea is to compress the data at the deployed nodes before it is transmitted to the sink where the original data is recovered with almost no loss. Thus, the major energy consuming task is shifted from the nodes to the sink. However the data reconstruction problem is bounded by mainly two issues: • The required number of measurements for reconstructing the signal.• The choice of the best suited algorithm for effective and accurate data recovery.Accurate and fast recovery of the original data at the sink is one of the major issues for most of the data recovery algorithms1. Existing approaches tend to have almost full recovery but it is at the cost of added computational complexity. Hence, an accptable tradeoff is seen between fast recovery of the original data and complexity of the recovery algorithms2,3. The Orthogonal Matching Pursuit (OMP) 4 algorithm is one of the most promising work in this field. The basic idea behind OMP is trivial and spontaneous; i.e. to choose a column of measurement matrix Φ in every iteration whose correlation is maximum with the residual. The key for the selected column of the measurement matrix, is appended to the list. Also, the remnant of the selected columns is removed from the Φ, updating a residual for the further iterations.