You are here

CrystalClear

Project researchers

CRYSTALCLEAR - Evaluating Synchronous Transmission Reliability under WiFi Interference in WiSHFUL

Co-existence issues may become a show-stopper for low-power wireless networks applied under the umbrella of the Internet of Things. A number of research activities have recently shown that omnipresent Wi-Fi devices may block data delivery in networks based on 2.4GHz IEEE 802.15.4 radios (so-called Low-Power and Lossy Networks, or LLNs). Several proposed techniques and protocol stacks aim at improving network dependability in the presence of external interference. Among them, Crystal is a protocol stack developed in our research group that provides support for sparse, aperiodic traffic with unprecedented high reliability, low latency and low energy consumption. The core technique underlying Crystal, i.e., synchronous transmissions (Glossy), provides significant advantages in terms of interference resilience, but cannot cope with strong and/or localized noise. Our current work enhances Crystal with mechanisms to efficiently survive interference by dynamically avoiding jammed frequencies and adapting the effort put toward packet delivery depending on the noise conditions. Our preliminary studies confirmed that the protocol delivers data fast and reliably in the presence of various kinds of interference patterns (Wi-Fi, microwave ovens) reproduced via JamLab on LLN devices. The use of JamLab is currently common practice in the LLN community, as it enables experiment reproducibility; however, the realism of the generated noise is questionable. In CrystalClear we will utilise the network control capabilities of WiSHFUL to generate real yet repeatable Wi-Fi interference, by exploiting the colocation of Wi-Fi and LLN devices in the TWIST testbed. This will enable the following research outcomes: 1) evaluation of Crystal (and competing protocols) against real yet repeatable interference 2) an in-field assessment of the limitations of emulated vs. real interference 3) the definition of a methodology for evaluating LLN devices, revolving around the hw/sw facilities in TWIST and their seamless availability via the WiSHFUL APIs and services.

Date: 
Friday, 1 December, 2017 to Monday, 30 April, 2018
Duration: 
5 months
Partners: 

Coordinator: Amy Lynn Murphy

Partners: UNITN

Funding: 
Fifth WiSHFUL Competitive Call for Experiments