As the title suggests, this book focuses on recent advances in the research involving compressed sensing for distributed systems. It is written by three researchers with expert knowledge in the field. Compressed sensing refers to the process of generating from a high-dimensional data observation lower-dimensional data that maintains the original information content. Such compression is possible because the data are assumed to have a high level of redundancy. Compressed sensing then follows a relatively traditional approach, generating the lower-dimensional result using linear transforms. While such transforms cannot be uniquely identified, adding additional constraints (such as orthogonality, or statistical independence) on the transform vectors results in methods that converge to unique solutions.
The novelty of the book does not rely on the restatement of compressed sensing and solutions, but on consideration of how compressed sensing works in a distributed system environment such as a sensor network. Here, additional restrictions (such as energy conservation, location, and availability of communication channels) make collaboration to extract the solution more complex. With only 100 pages, one would expect the book to be a quick read; however, the authors manage to provide rigorous and detailed coverage of the topics. The book is structured in four major chapters: a general introduction of the problems and three chapters focused on specific aspects such as rate distortion and joint and distributed recovery of the data. Each chapter is written as a standalone scientific paper, yet it assumes the reader to be familiar with the general problem concepts and have a good understanding of current algorithms used in information theory optimization problems. With that in mind, one would hope that the content of the book has been carefully reviewed by peers or has been vetted in other peer-reviewed publishing venues to ensure correctness.
The book provides an excellent overview of the state-of-the-art solutions addressing various situations such as collaboration of nodes versus independence of nodes or including energy conservation as a goal in the search for the solution. Moreover, the presented algorithms include discussions on their computational costs and are also presented with adequate supporting experimental results. Overall, the book is a highly technical, yet well-written contribution to the field of distributed systems and computational sensing and will find its audience among professionals (researchers, faculty, and graduate students) working in these fields.