Change search
ReferencesLink to record
Permanent link

Direct link
Interference Aware Self-Organization for Wireless Sensor Networks: a Reinforcement Learning Approach
KTH, School of Information and Communication Technology (ICT), Communication Systems, CoS.
KTH, School of Information and Communication Technology (ICT), Communication Systems, CoS.ORCID iD: 0000-0003-4986-6123
2008 (English)In: 2008 IEEE INTERNATIONAL CONFERENCE ON AUTOMATION SCIENCE AND ENGINEERING, VOLS 1 AND 2, NEW YORK: IEEE , 2008, 560-565 p.Conference paper (Refereed)
Abstract [en]

Reliability is a key issue in wireless sensor networks. Depending on the targeted application, reliability is achieved by establishing and maintaining a certain number of network functionalities: the greatest among those is certainly the capability of nodes to communicate. Sensors communications are sensible to interference that might corrupt packets transmission and even preclude the process of network formation. In this paper we propose a new scheme that allows to establish and maintain a connected topology while dealing with this problem. The idea of channel surfing (already introduced in [1]) is exploited to avoid interference; in the resulting multi-channel environment nodes discover their neighbors in a distributed fashion using a reinforcement learning (RL) algorithm. Our scheme allows the process of network formation even in presence of interference, overcoming thus the limit of algorithms currently implemented in state of the art standards for wireless sensor networks. By means of reinforcement learning the process of neighbor discovery is carried out in a fast and energy efficient way.

Place, publisher, year, edition, pages
NEW YORK: IEEE , 2008. 560-565 p.
Keyword [en]
Arts computing, Computer networks, Energy efficiency, Hybrid sensors, Learning algorithms, Network protocols, Reinforcement, Reinforcement learning, Reliability, Sensor networks, Sensors
National Category
URN: urn:nbn:se:kth:diva-25788DOI: 10.1109/COASE.2008.4626424ISI: 000261320400092ScopusID: 2-s2.0-54949119207ISBN: 978-1-4244-2022-3OAI: diva2:359881
IEEE International Conference on Automation Science and Engineering Arlington, VA, AUG 23-26, 2008 IEEE
QC 20101101Available from: 2010-11-01 Created: 2010-11-01 Last updated: 2010-11-01Bibliographically approved
In thesis
1. Design of Reliable Communication Solutions for Wireless Sensor Networks: Managing Interference in Unlicensed Bands
Open this publication in new window or tab >>Design of Reliable Communication Solutions for Wireless Sensor Networks: Managing Interference in Unlicensed Bands
2009 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Recent surveys conducted in the context of industrial automation have outlined that reliability concerns represent today one of the major barriers to the diffusion of wireless communications for sensing and control applications: this limits the potential of wireless sensor networks and slows down the adoption of this new technology. Overcoming these limitations requires that awareness on the causes of unreliability and on the possible solutions to this problem is created. With this respect, the main factor responsible for the perceived unreliability is radio interference: low-power communications of sensor nodes are in fact very sensitive to bad channel conditions and can be easily corrupted by transmissions of other co-located devices. In this thesis we investigate different techniques that can be exploited to avoid interference or mitigate its effects.We first consider interference avoidance through dynamic spectrum access: more specifically we focus on the idea of channel surfing and design algorithms that allow sensor nodes to identify interfered channels, discover their neighbors and maintain a connected topology in multi-channel environments. Our investigation shows that detecting and thus avoiding interference is a feasible task that can be performed by complexity and power constrained devices. In the context of spectrum sharing, we further consider the case of networked estimation and aim at quantifying the effects of intranetwork interference, induced by contention-based medium access, over the performance of an estimation system. We show that by choosing in an opportune manner their probability of transmitting, sensors belonging to a networked control system can minimize the average distortion of state estimates.In the second part of this thesis we focus on frequency hopping techniques and propose a new adaptive hopping algorithm. This implements a new approach for frequency hopping: in particular rather than aiming at removing bad channels from the adopted hopset our algorithm uses all the available frequencies but with probabilities that depend on the experienced channel conditions. Our performance evaluation shows that this approach outperforms traditional frequency hopping schemes as well as the adaptive implementation included in the IEEE 802.15.1 radio standard leading to a lower packet error rate.Finally, we consider the problem of sensor networks reprogramming and propose a way for ingineering a coding solution based on fountain codes and suitable for this challenging task. Using an original genetic approach we optimize the degree distribution of the used codes so as to achieve both low overhead and low decoding complexity. We further engineer the implementation of fountain codes in order to allow the recovery of corrupted information through overhearing and improve the resilience of the considered reprogramming protocol to channel errors.

Place, publisher, year, edition, pages
Stockholm: KTH, 2009. vii, 142 p.
Trita-ICT-COS, ISSN 1653-6347 ; 0901
National Category
urn:nbn:se:kth:diva-10257 (URN)
2009-05-14, Sal/Hall C1, KTH Electrum, Isafjordsgatan 26, KISTA, 14:00 (English)
Available from: 2009-05-07 Created: 2009-04-23 Last updated: 2010-11-01Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full textScopus

Search in DiVA

By author/editor
Stabellini, LucaZander, Jens
By organisation
Communication Systems, CoS

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 28 hits
ReferencesLink to record
Permanent link

Direct link