Lossy Compression for MANET Distributed Network Control

Distributed wireless networks are a viable approach for achieving communication in infrastructure-less scenarios such as D2D (device-to-device) communication, and MANET (Mobile Ad-Hoc Networks). In most cases, achieving communication with-out an infrastructure comes at a cost of control overhead. Control messages in MANETs grow proportionally with the number of nodes in the vicinity of the originator. Networks are becoming denser and larger, and the overhead incurred is limiting the possible growth and efficiency of current networks. In this work, we investigate different approaches for control message compression. Methods range from lossless compression, lossy compression using JPEG, and "directional" lossy compression using Bloom filter. The trade-off between correctness and compression is shown through comparative simulations. It is shown that while lossy compression presents high risks for information corruption, the ability to adjust the level and direction of errors is sufficient for gaining compression value along with a minimal impact on network activity.