Performance Limitations of Parsing Libraries: State-of-the-Art and Future Perspectives

The acceleration of mobile data traffic and the shortage of available spectral resources create new challenges for the next-generation (5G) networks. One of the potential solutions is network offloading that opens a possibility for unlicensed spectrum utilization. Heterogeneous networking between cellular and WLAN systems allows mobile users to adaptively utilize the licensed (LTE) and unlicensed (IEEE 802.11) radio technologies simultaneously. At the same time, softwarized frameworks can be employed not only inside the network controllers but also at the end nodes. To operate with the corresponding policies and interpret them efficiently, a signaling processor has to be developed and equipped with a fast packet parsing mechanism. In this scenario, the reaction time becomes a crucial factor, and this paper provides an overview of the existing parsing libraries (Scapy and dpkt) as well as proposes a flexible parsing tool that is capable of reducing the latency incurred by analyzing packets in a softwarized network.