Efficient management of data models in constrained systems by using templates and context based compression

Abstract

Data communication is at the heart of any distributed system. The adoption of generic data formats such as XML or JSON eases the exchange of information and interoperability among heterogeneous systems. However, the verbosity of those generic data formats usually requires system resources that might not be available in resource-constrained systems, e.g., embedded systems and those devices which are being integrated into the so-called IoT. In this work we present a method to reduce the cost of managing data models like XML or JSON by using templates and context based compression. We also provide a brief evaluation and comparison as a benchmark with current implementations of W3C’s Efficient XML Interchange (EXI) processor. Although the method described in this paper is still at its initial stage, it outperforms the EXI implementations in terms of memory usage and speed, while keeping similar compression rates. As a consequence, we believe that our approach fits better for constrained systems.