Time Slotted Channel Hopping (TSCH) is an amendment of the the IEEE 802.15.4 working group to provide a low-power Medium Access Control (MAC) for the Internet of Things (IoT). This standard relies on techniques such as channel hopping and bandwidth reservation to ensure both energy savings and reliable transmissions. Since many applications require low end-to-end delay (e.g. alarms), we propose here a distributed algorithm to schedule the transmissions with a short end-to-end delay. We divide the network in stratums, regrouping all the nodes with the same depth in the DODAG constructed by RPL. Then, different time-frequency blocks are assigned deterministically to each stratum. By appropriately organizing the blocks in the slotframe, we are able to deliver a packet before the end of the slotframe, whatever the route length is. We present a simple analytical study to define the initial size of each block in a homogeneous scenario. We experimentally analyze the behavior of our strategy to validate its ability to provide both high reliability and low latency in a distributed manner.

The Industrial Internet of Things tends now to emerge as a key paradigm to interconnect a collection of wireless devices. However, most industrial applications have strict requirements, especially concerning the reliability and the latency. IEEE802.15.4-TSCH represents currently a promising standard relying on a strict schedule of the transmissions to provide such guarantees. The standard ISA-100.11a-2011 has proposed the concept of duocast, where a pair of receivers are allocated to the same transmission opportunity to increase the reliability. In this paper, we generalize this approach to involve k different receivers, and we explore the impact of this technique on the performance of the network. We propose an algorithm assigning several receivers for each transmission to increase the probability that at least one device receives correctly the packet. By exploiting a multipath topology created by the routing layer, we are able to reduce the number of transmissions while still achieving the same reliability. We consequently increase the network capacity, and reduce significantly the jitter. Our simulation results highlight the relevance of this k-cast technique in TSCH for the Industrial Internet of Things.

The IoT expects to exploit IEEE802.15.4e-TSCH, designed for wireless industrial sensor networks. This standard relies on techniques such as channel hopping and bandwidth reservation to ensure both energy savings and reliable transmissions. The 6TiSCH working group currently proposes to exploit the RPL routing protocol on top of the IEEE802.15.4-2012-TSCH layer. Since many applications may require low end-to-end delay (e.g. alarms), we propose here a distributed algorithm to schedule the transmissions while upper bounding the end-to-end delay. Our strategy is based on stratums to reserve time-bands for each depth in the routing structure constructed by RPL. By allocating a sufficient number of timeslots for the possible retransmissions, we guarantee that any packet is delivered during one single slotframe, wherever the source is located. Experiments on a large scale testbed prove the relevance of this approach to reduce the end-to-end delay while minimizing the number of collisions, prejudicial to the reliability in multihop networks.

No abstract

No abstract

This paper focuses on the issue of distributed reinforcement learning (RL) for decision-making in cooperative multi-agent systems. Although this problem has been a topic of interest to many researchers, results obtained from these works aren't sufficient and several difficulties have not yet been resolved, such as, the curse of dimensionality and the multi-agent coordination problem. These issues are aggravated exponentially as the number of agents or states increases, resulting in large memory requirement, slowness in learning speed, coordination failure and even no system convergence. As a solution, a new distributed RL algorithm, called the ThMLA-JAG method, is proposed here. Its main idea is to decompose the coordination of all agents into several two-agent coordination and to use a teammate model for managing other agents' experiences. Validation tests on a pursuit game show that the proposed method overcomes the aforementioned limitations and is a good alternative to RL methods when dealing with cooperative learning in dynamics environments while avoiding collisions with obstacles and other learners.

Abstract:Communications infrastructure is a basic part for the success of smart grid. Optimisation of energy consumption in the future intelligent energy networks (smart grids) will be based on the integrated near real-time between the different elements of grid network communications. Through a communication infrastructure, a smart grid can improve the reliability of power, eliminate power outages, and optimise energy consumption. The network gets smarter by improving detection, so the most important element that makes the smart grid is wireless sensor networks. Among them, one of the most important applications is a sensor data collection. This paper discusses some of communications research challenges and opportunities in the areas of smart grid and smart meters. In particular, we focus on some of the main communications challenges to achieve interoperability and future proof, smart grid/metering networks. We describe the basic taxonomy and propose to break the network data collection of wireless sensors.

No abstract