Performance Enhancement Proxies (PEPs) are widely used to improve TCP performance in geostationary orbit (GEO) satellite networks, which experience long RTT (approximately 500 ms). As a split TCP-based solution, PEP divides the end-to-end connection into multiple sub-connections, each independently managing its own rate control, including both congestion and flow control. This independence naturally leads to rate imbalances, manifested as an abnormal on-off traffic pattern-a phenomenon confirmed by our experimental observations in real-world GEO satellite networks. Furthermore, we demonstrate that this on-off traffic pattern not only undermines fairness but also reduces throughput, particularly for small-sized flows. To address this problem, we propose PEP-Policer, an automatic rate limiter for PEP to limit the link with higher rate to the lower one. Unlike traditional rate limiting methods that require manual configuration of the target rate, PEP-Policer employs a finite state machine to automatically determine and enforce the appropriate target rate. Moreover, as an add-on, PEP-Policer is compatible with all PEP-based solutions and can be integrated without modifying existing systems. Extensive evaluations in an emulated GEO satellite network show that PEP-Policer improves TCP's fairness and increases goodput by up to 63.0% for Cubic, 49.6% for BBR, and 88.6% for Hybla.

The Satellite-Terrestrial Integrated Network (STIN) enhances end-to-end transmission by simultaneously utilizing terrestrial and satellite networks, offering significant benefits in scenarios like emergency response and cross-continental communication. Low Earth Orbit (LEO) satellite networks offer reduced Round Trip Time (RTT) for long-distance data transmission and serve as a crucial backup during terrestrial network failures. Meanwhile, terrestrial networks are characterized by ample bandwidth resources and generally more stable link conditions. Therefore, integrating Multipath TCP (MPTCP) into STIN is vital for optimizing resource utilization and ensuring efficient data transfer by exploiting the complementary strengths of both networks. However, the inherent challenges of STIN, such as heterogeneity, instability, and handovers, pose difficulties for traditional multipath schedulers, which are typically designed for terrestrial networks. We propose a novel multipath data scheduling approach for STIN, the Adaptive Latency Compensation Scheduler (ALCS), to address these issues. ALCS refines transmission latency estimates by incorporating RTT, congestion window size, inflight and queuing packets, and satellite trajectory information. It further employs adaptive mechanisms for latency compensation and proactive handover management. Implemented in the MPTCP Linux Kernel and evaluated in a simulated STIN testbed, ALCS outperforms existing multipath schedulers, delivering faster data transmission and achieving throughput gains of 9.8% to 44.0% compared to benchmark algorithms.

The construction of Low Earth Orbit satellite constellations has recently spurred tremendous attention from both academia and industry. 5G and 6G standards have specified the LEO satellite network as a key component of the mobile network. However, due to the satellites' fast traveling speed, ground terminals usually experience frequent and high-latency handover, which significantly deteriorates the performance of latencysensitive applications. To address this challenge, we propose a parallel handover mechanism for the mobile satellite network which can considerably reduce the handover latency. The main idea is to use plan-based handovers instead of measurementbased handovers to avoid interactions between the access and core networks, hence eliminating the significant time overhead in the traditional handover procedure. Specifically, we introduce a novel network function named Satellite Synchronized Function (SSF), which is designed for being compliant with the standard 5G core network. Moreover, we propose a machine learning model for signal strength prediction, coupled with an efficient handover scheduling algorithm. We have conducted extensive experiments and results demonstrate that our proposed handover scheme can considerably reduce the handover latency by 21× compared to the standard NTN handover scheme and two other existing handover schemes, along with significant improvements in network stability and user-level performance.

Wireless federated learning (WFL) suffers from heterogeneity prevailing in the data distributions, computing powers, and channel conditions of participating devices. This paper presents a new Federated Learning with Adjusted leaRning ratE (FLARE) framework to mitigate the impact of the heterogeneity. The key idea is to allow the participating devices to adjust their individual learning rates and local training iterations, adapting to their instantaneous computing powers. The convergence upper bound of FLARE is established rigorously under a general setting with non-convex models in the presence of non-i.i.d. datasets and imbalanced computing powers. By minimizing the upper bound, we further optimize the scheduling of FLARE to exploit the channel heterogeneity. A nested problem structure is revealed to facilitate iteratively allocating the bandwidth with binary search and selecting devices with a new greedy method. A linear problem structure is also identified and a low-complexity linear programming scheduling policy is designed when training models have large Lipschitz constants. Experiments demonstrate that FLARE consistently outperforms the baselines in test accuracy, and converges much faster with the proposed scheduling policy.

State machine replication (SMR) uses consensus as its core component for reaching agreement among a group of processes, in order to provide fault-tolerant services. Most SMR protocols, such as Paxos and Raft, are designed in the partial synchrony model. Partially synchronous protocols rely on timing assumptions to elect a special role (such as the leader), which may become the performance bottleneck under a heavy workload. From an engineering perspective, partially synchronous protocols have to wait for a pre-defined period of time and implement a (complicated) failover mechanism in order to replace the faulty leader. In contrast, asynchronous protocols are immune to such problems.

The construction of Low Earth Orbit (LEO) satellite constellations has recently spurred tremendous attention from academia and industry. 5G and 6G standards have specified LEO satellite network as a key component of 5G and 6G networks. However, ground terminals experience frequent, high-latency handover incurred by satellites’ fast travelling speed, which deteriorates the performance of latency-sensitive applications. To address this challenge, we propose a novel handover flowchart for mobile satellite networks, which can considerably reduce the handover latency. The innovation behind this scheme is to mitigate the interaction between the access and core networks that occupy the majority of time overhead by leveraging the predictable travelling trajectory and spatial distribution inherent in mobile satellite networks. Specifically, we design a fine-grained synchronized algorithm to address the synchronization problem due to the lack of control signalling delivery between the access and core networks. Moreover, we minimize the computational complexity of the core network using information such as the satellite access strategy and unique spatial distribution, which is caused by frequent prediction operations. We have built a prototype for a mobile satellite network using modified Open5GS and UERANSIM, which is driven by actual LEO satellite constellations such as Starlink and Kuiper. We have conducted extensive experiments, and the results demonstrate that our proposed handover scheme can considerably reduce the handover latency compared to the 3GPP Non-terrestrial Networks (NTN) and two other existing handover schemes.

Asynchronous Byzantine Fault-Tolerant (BFT) consensus protocols maintain strong consistency across nodes (i.e., ensure safety) and terminate probabilistically (i.e., ensure liveness) despite unbounded network delay. In contrast to protocols under partial synchrony, asynchronous counterparts pay no extra timing assumptions for electing a special role, and thus is more robust to network issues. To formally study this feature, we propose a new classification method for consensus and accordingly categorize relevant work: timing-balanced protocols are those that do not introduce strictly stronger timing-related assumptions for liveness, compared to ones required by safety.

Wall-clock convergence time and communication load are key performance metrics for the distributed implementation of stochastic gradient descent (SGD) in parameter server settings. Communication-adaptive distributed Adam (CADA) has been recently proposed as a way to reduce communication load via the adaptive selection of workers. CADA is subject to performance degradation in terms of wall-clock convergence time in the presence of stragglers. This paper proposes a novel scheme named grouping-based CADA (G-CADA) that retains the advantages of CADA in reducing the communication load, while increasing the robustness to stragglers at the cost of additional storage at the workers. G-CADA partitions the workers into groups of workers that are assigned the same data shards. Groups are scheduled adaptively at each iteration, and the server only waits for the fastest worker in each selected group. We provide analysis and experimental results to elaborate the significant gains on the wall-clock time, as well as communication load and computation load, of G-CADA over other benchmark schemes.

Consider a multi-cell mobile edge computing network, in which each user wishes to compute the product of a user-generated data matrix with a network-stored matrix. This is done through task offloading by means of input uploading, distributed computing at edge nodes (ENs), and output downloading. Task offloading may suffer long delay since servers at some ENs may be straggling due to random computation time, and wireless channels may experience severe fading and interference. This paper aims to investigate the interplay among upload, computation, and download latencies during the offloading process in the high signal-to-noise ratio regime from an information-theoretic perspective. A policy based on cascaded coded computing and on coordinated and cooperative interference management in uplink and downlink is proposed and proved to be approximately optimal for a sufficiently large upload time. By investing more time in uplink transmission, the policy creates data redundancy at the ENs, which can reduce the computation time, by enabling the use of coded computing, as well as the download time via transmitter cooperation. Moreover, the policy allows computation time to be traded for download time. Numerical examples demonstrate that the proposed policy can improve over existing schemes by significantly reducing the end-to-end execution time.

Gradient-based distributed learning in parameter server (PS) computing architectures is subject to random delays due to straggling worker nodes and to possible communication bottlenecks between PS and workers. Solutions have been recently proposed to separately address these impairments based on the ideas of gradient coding (GC), worker grouping, and adaptive worker selection. This article provides a unified analysis of these techniques in terms of wall-clock time, communication, and computation complexity measures. Furthermore, in order to combine the benefits of GC and grouping in terms of robustness to stragglers with the communication and computation load gains of adaptive selection, novel strategies, named lazily aggregated GC (LAGC) and grouped-LAG (G-LAG), are introduced. Analysis and results show that G-LAG provides the best wall-clock time and communication performance while maintaining a low computational cost, for two representative distributions of the computing times of the worker nodes.

A multi-cell mobile edge computing network is studied, in which each user wishes to compute the product of a user-generated data matrix with a network-stored matrix through data uploading, distributed edge computing, and output downloading. Assuming randomly straggling edge servers, this paper investigates the interplay among upload, compute, and download times in high signal-to-noise ratio regimes. A policy based on cascaded coded computing and on coordinated and cooperative interference management in uplink and downlink is proposed and proved to be approximately optimal for sufficiently large upload times. By investing more time in uplink transmission, the policy creates data redundancy at the edge nodes to reduce both computation times by coded computing, and download times via transmitter cooperation. Moreover, it allows computing times to be traded for download times.

This work identifies the optimal Degrees-ofFreedom (DoF) of the K-User MISO Broadcast Channel (BC) with delayed Channel-State Information at the Transmitter (CSIT) and with additional current noisy CSIT where the current channel estimation error scales in P-αfor α ∈ [0, 1]. These two settings had in the past been studied separately; the setting of imperfect current CSIT has attracted considerable interest over the last decade, while the setting of delayed CSIT was studied in the seminal work of Maddah-Ali and Tse in 2010 1 where an optimal DoF of K/ Σk=1K1/k was established. Since k=1 then there have been several efforts to combine the two settings of delayed and imperfect-current CSIT. Our work establishes for the first time the optimal DoF in this joint setting, capitalizing on a novel transmission scheme that is presented here, which combines a structurally new approach of handling past and current interference, to achieve the optimal performance. We establish the once elusive optimal DoF to be of the form 1 αK + (1 - α)K/(K/ Σk=1K1/k). This further shows that the two k=1 types of DoF gains, from current and delayed CSIT, can be combined additively.

In a distributed computing system operating according to the map-shuffle-reduce framework, coding data prior to storage can be useful both to reduce the latency caused by straggling servers and to decrease the inter-server communication load in the shuffle phase. In prior work, a concatenated coding scheme was proposed for a matrix multiplication task. In this scheme, the outer Maximum Distance Separable (MDS) code is leveraged to correct erasures caused by stragglers, while the inner repetition code is used to improve the communication efficiency in the shuffle phase by means of coded multi-casting. In this work, it is demonstrated that it is possible to leverage the redundancy created by repetition coding in order to increase the rate of the outer MDS code and hence to increase the multicasting opportunities in the shuffle phase. As a result, the proposed approach is shown to improve over the best known latency-communication overhead trade-off.

Distributed computing platforms typically assume the availability of reliable and dedicated connections among the processors. This work considers an alternative scenario, relevant for wireless data centers and federated learning, in which the distributed processors, operating on generally distinct coded data, are connected via shared wireless channels accessed via full-duplex transmission. The study accounts for both wireless and computing impairments, including interference, imperfect Channel State Information, and straggling processors, and it assumes a Map-Shuffle-Reduce coded computing paradigm. The total latency of the system, obtained as the sum of computing and communication delays, is studied for different shuffling strategies revealing the interplay between distributed computing, coding, and cooperative or coordinated transmission.

This paper considers a cloud-RAN architecture with cache-enabled multi-antenna Edge Nodes (ENs) that deliver content to cache-enabled end-users. The ENs are connected to a central server via limited-capacity fronthaul links, and, based on the information received from the central server and the cached contents, they transmit on the shared wireless medium to satisfy users' requests. By leveraging cooperative transmission as enabled by ENs' caches and fronthaul links, as well as multicasting opportunities provided by users' caches, a close-to-optimal caching and delivery scheme is proposed. As a result, the minimum Normalized Delivery Time (NDT), a high-SNR measure of delivery latency, is characterized to within a multiplicative constant gap of 3/2 under the assumption of uncoded caching and fronthaul transmission, and of one-shot linear precoding. This result demonstrates the interplay among fronthaul links capacity, ENs' caches, and end-users' caches in minimizing the content delivery time.

In fog-aided cellular systems, content delivery latency can be minimized by jointly optimizing edge caching and transmission strategies. In order to account for the cache capacity limitations at the edge nodes (ENs), transmission generally involves both fronthaul transfer from a cloud processor with access to the content library to the ENs and wireless delivery from the ENs to the users. In this paper, the resulting problem is studied from an information-theoretic viewpoint by making the following practically relevant assumptions: 1) the ENs have multiple antennas; 2) only uncoded fractional caching is allowed; 3) the fronthaul links are used to send fractions of contents; and 4) the ENs are constrained to use one-shot linear zero-forcing precoding on the wireless channel. Assuming off-line proactive caching and focusing on a high signal-to-noise ratio (SNR) latency metric, the optimal information-theoretic performance is investigated under both serial and pipelined fronthaul-edge transmission modes. The analysis characterizes the minimum high-SNR latency in terms of normalized delivery time (NDT) for worst case users' demands. The characterization is exact for a subset of system parameters and is generally optimal within a multiplicative factor of 3/2 for the serial case and 2 for the pipelined case. The results bring insights into the optimal interplay between edge and cloud processing in fog-aided wireless networks as a function of system resources, including the number of antennas at the ENs, the ENs' cache capacity, and the fronthaul capacity.

In cellular systems, content delivery latency can be minimized by jointly optimizing edge caching, fronthaul transmission from a cloud processor (CP) with access to the content library, and wireless transmission. In this paper, this problem is studied from an information-theoretic viewpoint by making the following practically relevant assumptions: 1) the ENs have multiple antennas; 2) only uncoded fractional caching is allowed; 3) the fronthau llinks are used to send fractions of contents; and 4) the ENs are constrained to use one-shot linear precoding on the wireless channel. Assuming offline caching and focusing on a high signal-to-noise ratio (SNR) latency performance metric, the proposed caching and delivery policy is shown to be either exactly optimal or optimal within a multiplicative factor of 3/2. The results bring insights into the optimal interplay between edge and cloud processing in fog-aided wireless networks as a function of system resources, including the number of antennas at the ENs, the ENs' cache capacity and the fronthaul capacity.

In a Fog Radio Access Network (F-RAN), the cloud processor (CP) collects channel state information (CSI) from the edge nodes (ENs) over fronthaul links. As a result, the CSI at the cloud is generally affected by an error due to outdating. In this work, the problem of content delivery based on fronthaul transmission and edge caching is studied from an information-theoretic perspective in the high signal-to-noise ratio (SNR) regime. For the set-up under study, under the assumption of perfect CSI, prior work has shown the (approximate or exact) optimality of a scheme in which the ENs transmit information received from the cloud and cached contents over orthogonal resources. In this work, it is demonstrated that a non-orthogonal transmission scheme is able to substantially improve the latency performance in the presence of imperfect CSI at the cloud.

International audience

This work explores cache-aided interference management in the absence of channel state information at the transmitters (CSIT), focusing on the setting with K transmitter/receiver pairs endowed with caches, where each receiver k is connected to transmitter k via a direct link with normalized capacity 1, and to any other transmitter via a cross link with normalized capacity τ ≤ 1. In this setting, we explore how a combination of pre-caching at transmitters and receivers, together with interference enhancement techniques, can a) partially counter the lack of CSIT, and b) render the network self-sufficient, in the sense that the transmitters need not receive additional data after pre-caching. Toward this we present new schemes that blindly harness topology and transmitter-and-receiver caching, to create separate streams, each serving many receivers at a time. Key to the approach here is a combination of rate-splitting, interference enhancement and coded caching.

We explore the performance of coded caching in a SISO BC setting where some users have higher link capacities than others. Focusing on a binary and fixed topological model where strong links have a fixed normalized capacity 1, and where weak links have reduced normalized capacity Tgwhere g is the coded-caching gain, and where w is the fraction of users that are weak. This leads to the interesting conclusion that for coded multicasting, the weak users need not bring down the performance of all users, but on the contrary to a certain extent, the strong users can lift the performance of the weak users without any penalties on their own performance. Furthermore for smaller ranges of τ, we also see that achieving the near-optimal performance comes with the advantage that the strong users do not suffer any additional delays compared to the case where T = 1.

Building on the recent coded-caching breakthrough by Maddah-Ali and Niesen, the work here considers the $K$ -user cache-aided wireless multi-antenna symmetric broadcast channel with random fading and imperfect feedback, and analyzes the throughput performance as a function of feedback statistics and cache size. In this setting, this paper identifies the optimal cache-aided degrees-of-freedom (DoF) within a factor of 4, by identifying near-optimal schemes that exploit a new synergy between coded caching and delayed CSIT, as well as by exploiting the unexplored interplay between caching and feedback-quality. The DoF expressions reveal an initial gain due to current CSIT, and an additional gain due to coded caching, which is exponential in the sense that any linear decrease in the required DoF performance, allows for an exponential reduction in the required cache size. In the end, this paper reveals three new aspects of caching: a synergy between memory and delayed feedback, a tradeoff between memory and current CSIT, and a powerful ability to provide cache-aided feedback savings.

This work1derives the optimal Degrees-of-Freedom (DoF) of the K-User MISO Broadcast Channel (BC) with delayed Channel-State Information at the Transmitter (CSIT) and with additional current noisy CSIT where the channel estimation error scales in P-αfor α ∈ [0, 1]. The optimal sum DoF takes the simple form (1 - α)K/HK+ αK where HK =△ Σk=1K1/k. This optimal performance is the result of a novel scheme which deviates from existing efforts as it digitally combines interference, decodes symbols of any order in the MAT alignment [1], and utilizes a hierarchical quantizer whose output is distributed across rounds in a way that minimizes unwanted interference. These jointly deliver, for the first time, the elusive DoF-optimal combining of MAT and ZF.

In the setting of the two-user single-input single-output X channel, recent works have explored the degrees-offreedom (DoF) limits in the presence of perfect channel state information at the transmitter (CSIT), as well as in the presence of perfect-quality delayed CSIT. Our work shows that the same DoF-optimal performance - previously associated to perfect-quality current CSIT - can in fact be achieved with current CSIT that is of imperfect quality. The work also shows that the DoF performance previously associated to perfect-quality delayed CSIT, can in fact be achieved in the presence of imperfect-quality delayed CSIT. These follow from the presented sum-DoF lower bound that bridges the gap - as a function of the quality of delayed CSIT - between the cases of having no feedback and having delayed feedback, and then another bound that bridges the DoF gap - as a function of the quality of current CSIT - between delayed and perfect current CSIT. The bounds are based on novel precoding schemes that are presented here and which employ imperfect-quality current and/or delayed feedback to align interference in space and in time.