Federated learning (FL) is a machine learning (ML) technique that enables collaborative model training without sharing raw data, making it ideal for Internet of Things (IoT) applications where data are distributed across devices and privacy is a concern. Wireless Sensor Networks (WSNs) play a crucial role in IoT systems by collecting data from the physical environment. This paper presents a comprehensive survey of the integration of FL, IoT, and WSNs. It covers FL basics, strategies, and types and discusses the integration of FL, IoT, and WSNs in various domains. The paper addresses challenges related to heterogeneity in FL and summarizes state-of-the-art research in this area. It also explores security and privacy considerations and performance evaluation methodologies. The paper outlines the latest achievements and potential research directions in FL, IoT, and WSNs and emphasizes the significance of the surveyed topics within the context of current technological advancements.

In the Internet of Things (IoT), a wireless sensor network (WSN) is deployed for collecting the interesting data of an application field. Sensor nodes in an IoT WSN are usually with the heterogeneous property. Some nodes have more power (energy) and additional functionality (e.g., data aggregation) than others. Cluster-based routing is usually used in WSNs for data transmissions due to efficiently routing consideration. Based on cluster-based routing, the cluster heads (CHs) act as the sensed data forwarding role. Once one or more CHs fail, the faulty CHs cannot forward the sensed data of their serving sensor nodes. As a result, the sink node (gateway) has not sufficient sensed data of the IoT application field. This will deeply affect the information processing of the IoT applications. We utilize the virtual CH formation and flow graph modeling to efficiently tolerate the failures of CHs. First, the available resources of all failure-free CHs are logically organized as a virtual CH to be the common backup of all faulty CHs. Then, the flow graph modeling is used to achieve fault tolerance with the minimum total energy consumption among all failure-free CHs. Finally, we perform extensive experiments to demonstrate the effectiveness of our approach in the fault-tolerant routing of the IoT WSNs.

Randomized scheduling is a well-known scheduling algorithm for wireless sensor networks to prolong network lifetime. In this algorithm, sensor nodes are randomly assigned to multiple working subsets. These subsets alternatively perform the sensing tasks for monitoring a sensor filed. In each subset, only a portion of sensor nodes are active. It is difficult to maintain the full sensing coverage and network connectivity in a sensor field. For the network connectivity problem of the randomized scheduling, it has been discussed. In this paper, we propose a distributed approach for assisting the randomized scheduling to provide the full coverage. To efficiently achieve this goal, the proposed approach is divided into two main stages: field partition and coverage improvement. The handling problems of these two stages can be transferred to two geometry problems: Voronoi polygon construction and circle covering. Simulation results show that the proposed approach can significantly improve the coverage quality of the randomized scheduling.

Edge computing is a promising alternative to cloud computing for offloading computationally heavy tasks from resource-constrained mobile user devices. Placed at the edge of the network, edge computing is particularly advantageous to delay-limited applications for having a short distance to end-users. However, when a mobile user moves away from the service coverage of the associated edge server, the advantage gradually vanishes, increasing response time. Although service migration has been studied to address this problem focusing on minimizing the service downtime, both zero-downtime and the amount of traffic generated as a result of migration need further study. In this paper, an optimal live migration for containerized edge computing service is studied. This paper presents three zero-downtime migration techniques based on state duplication and state reproduction techniques, and then, proposes an optimal migration technique selection algorithm that jointly minimizes the response time and network traffic during migration. For validation and performance comparison, the proposed migration techniques are implemented on off-the-shelf hardware with Linux operating system. The evaluation results showed that compared with a naive migration, the optimal approach reduced the response time and network load by at least 74.75% and 94.79%, respectively, under considered scenarios.