Distributed Sparse Random Projections for Refinable Approximation

Consider a large-scale wireless sensor network measuring compressible data, where n distributed data values can be well-approximated using only k