To leverage massive distributed data and computation resources, machine learning in the network edge is considered to be a promising technique, especially for large-scale model training. Federated learning (FL), as a paradigm of collaborative learning techniques, has obtained increasing research attention with the benefits of communication efficiency and improved data privacy. Due to the lossy communication channels and limited communication resources (e.g., bandwidth and power), it is of interest to investigate fast responding and accurate FL schemes over wireless systems. Hence, we investigate the problem of jointly optimized communication efficiency and resources for FL over wireless Internet of Things (IoT) networks. To reduce complexity, we divide the overall optimization problem into two subproblems, i.e., the client scheduling problem and the resource allocation problem. To reduce the communication costs for FL in wireless IoT networks, a new client scheduling policy is proposed by reusing stale local model parameters. To maximize successful information exchange over networks, a Lagrange multiplier method is first leveraged by decoupling variables, including power variables, bandwidth variables, and transmission indicators. Then, a linear-search-based power and bandwidth allocation method is developed. Given appropriate hyperparameters, we show that the proposed communication-efficient FL (CEFL) framework converges at a strong linear rate. Through extensive experiments, it is revealed that the proposed CEFL framework substantially boosts both the communication efficiency and learning performance of both training loss and test accuracy for FL over wireless IoT networks compared to a basic FL approach with uniform resource allocation.

Cell-free networks are considered to be a promising distributed network architecture to satisfy the increasing number of users and high rate expectations in beyond-5G systems. However, to further enhance network capacity, an increasing number of high-cost base stations (BSs) is required. To address this problem and inspired by the cost-effective intelligent reflecting surface (IRS) technique, we propose a fully decentralized design framework for cooperative beamforming in IRS-aided cell-free networks. We first transform the centralized weighted sum-rate maximization problem into a tractable consensus optimization problem, and then an incremental alternating direction method of multipliers (ADMM) algorithm is proposed to locally update the beamformer. The complexity and convergence of the proposed method are analyzed, and these results show that the performance of the new scheme can asymptotically approach that of the centralized one as the number of iterations increases. Results also show that IRSs can significantly increase the system sum-rate of cell-free networks and the proposed method outperforms existing decentralized methods.

A user wants to retrieve a file from a database without revealing the identity of the file retrieved to the operator of the database (server), which is known as the problem of private information retrieval (PIR). If it is further required that the user obtains no information about the other files in the database, the concept of symmetric private information retrieval (SPIR) is introduced to guarantee privacy for both parties. For SPIR, the server(s) need to access some randomness independent of the database, to protect the content of undesired files from the user. The information-theoretic capacity of SPIR is defined as the maximum number of information bits of the desired file retrieved per downloaded bit. In this paper, the problem of SPIR is studied for a distributed storage system with N servers (nodes), where all data (including the files and the randomness) are stored in a distributed way. Specifically, the files are stored by an (N,KC)-MDS storage code. The randomness is distributedly stored such that any KC servers store independent randomness information. We consider two scenarios regarding to the ability of the storage nodes to cooperate. In the first scenario considered, the storage nodes do not communicate or collude. It is shown that the SPIR capacity for MDS-coded storage (hence called MDS-SPIR) is 1-KCN, when the amount of the total randomness of distributed nodes (unavailable at the user) is at least KCN-KC times the file size. Otherwise, the MDS-SPIR capacity equals zero. The second scenario considered is the T-colluding SPIR problem (hence called TSPIR). Specifically, any T out of N servers may collude, that is, they may communicate their interactions with the user to guess the identity of the requested file. In the special case with KC=1, i.e., the database is replicated at each node, the capacity of TSPIR is shown to be 1-TN, with the ratio of the total randomness size relative to the file size be at least TN-T. For TSPIR with MDS-coded storage (called MDS-TSPIR for short), when restricted to schemes with additive randomness where the servers add the randomness to the answers regardless of the queries received, the capacity is proved to equal 1-KC+T-1N, with total randomness at least KC+T-1N-KC-T+1 times the file size. The MDS-TSPIR capacity for general schemes remains an open problem.

With the rise of critical machine-to-machine applications, next generation wireless communication systems must meet challenging requirements with respect to latency and reliability. A key question in this context relates to channel state estimation, which allows the transmitter to adapt the code rate to the channel state. In this work, we characterize the trade-off between the training sequence length and data codeword length: shorter channel estimation leaves more time for the payload transmission but reduces the estimation accuracy and causes more decoding errors. Using lower coding rates can mitigate this effect, but may result in a higher backlog of data at the transmitter. In order to optimize the training sequence length and the rate adaptation scheme with respect to the delay performance, we employ queueing analysis on top of accurate models of the physical layer. We obtain an analytically tractable solution to the problem by deriving a closed-form approximation for the decoding error probability due to imperfect channel knowledge and finite blocklength channel coding. The optimized training sequence length and rate adaptation strategy can reduce the delay violation probability by an order of magnitude, compared to suboptimal strategies that do not consider the delay constraints.

We study whether using non-orthogonal multiple access (NOMA) in the uplink of a mobile network can reduce the queueing delay compared to orthogonal multiple access (OMA) when the system requires communications at very low latency and high reliability. We first consider an ideal system model with perfect channel state information (CSI) at the transmitter and long codewords, where we determine the optimal decoding orders when the decoder uses successive interference cancellation (SIC) and derive closed-form expressions for the optimal rate when joint decoding is used. While joint decoding performs well even under tight delay constraints, NOMA with SIC decoding often performs worse than OMA. For low-latency systems, we must also consider the impact of finite-length channel coding, as well as rate adaptation based imperfect CSI. We derive closed-form approximations for the corresponding outage or error probabilities and find that those effects create a larger performance penalty for NOMA than for OMA. Thus, NOMA with SIC decoding may often be unsuitable for low-latency systems.

We consider the problem of private information retrieval (PIR) with colluding servers and eavesdroppers (abbreviated as ETPIR). The ETPIR problem is comprised of K messages, N servers where each server stores all K messages, a user who wants to retrieve one of the K messages without revealing the desired message index to any set of T colluding servers, and an eavesdropper who can listen to the queries and answers of any E servers but is prevented from learning any information about the messages. The information theoretic capacity of ETPIR is defined to be the maximum number of desired message symbols retrieved privately per information symbol downloaded. We show that the capacity of ETPIR is C = (1 – E/N) (1 + T–E/N–E + … + ( T–E/N–E) K–1)–1 when E < T, and C = (1 – E/N) when E ≥ T. To achieve the capacity, the servers need to share a common random variable (independent of the messages), and its size must be at least E/N · 1/C symbols per message symbol. Otherwise, with less amount of shared common randomness, ETPIR is not feasible and the capacity reduces to zero. An interesting observation is that the ETPIR capacity expression takes different forms in two regimes. When E < T, the capacity equals the inverse of a sum of a geometric series with K terms and decreases with K; this form is typical for capacity expressions of PIR. When E ≥ T, the capacity does not depend on K, a typical form for capacity expressions of SPIR (symmetric PIR, which further requires data-privacy, i.e., the user learns no information about other undesired messages); the capacity does not depend on T either. In addition, the ETPIR capacity result includes multiple previous PIR and SPIR capacity results as special cases.