Computer scientists usually view evolutionary computation as a sub-field of artificial intelligence involving combinatorial optimization. Evolutionary computation uses iterative processes, combined with random search (eventually parallelized), to meet some optimization goals.
The editors and authors of this collection of papers intend to show the relevance of evolutionary computation in a set of established energy-saving applications (data centers, wireless sensor networks, virtualized Web services, and grid computing). As a result, the relevance of the volume is limited to researchers in the same field, and to other researchers in green computing who are looking for approaches to solve optimization problems.
The volume comprises eight chapters devoted to selected high-performance computing and network subsystem energy-saving efforts. Because each application field has to be surveyed, and the corresponding model formulated, large parts of these chapters are devoted to contextual information. However, algorithmic contributions are found in chapter 3 on macro-level models for power consumption in servers, and in chapter 4 on genetic-based batch schedulers. Some computational results are reported in these chapters, as well as in chapter 5 on power consumption constrained scheduling; in chapter 6, about on-chip wireless communication for massive multicore systems; and in chapter 8, on data compression and node localization in wireless sensor networks.
Each chapter has its own list of references, but the volume lacks an index and a consolidated reference list. While meta-heuristics based on some high-level models are presented, it would have been useful to compare their results, when available, with the real heuristics or other algorithms as implemented in the systems on which experiments are reported. One reason is obviously to account for modeling uncertainty and the typical randomness in many energy-consuming processes.