Flying Ad Hoc Networks (FANETs) formed by Unmanned Aerial Vehicles (UAVs) with multiple heterogeneous radios are increasingly being adopted for various applications. However, the presence of asymmetric links caused by hardware differences poses challenges to common routing protocols designed for homogeneous FANETs. Ignoring asymmetric links may lead to inefficient resource utilization and compromise network connectivity, resulting in a significant decline in the performance of existing protocols. How to fully utilize the abundant link resources to enhance the network performance and offer more routing options for data with diversified service requirements, is a challenging task. In this paper, we propose an asymmetric link-based routing protocol in heterogeneous multi-radio FANETs, named ARHet. To efficiently exploit the asymmetric links, we first propose an asymmetric link discovery and information feedback mechanism. Then, to satisfy the specific service requirement of each type of data, we propose a traffic-differentiated routing strategy that can establish the appropriate paths for them. Experimental results show that ARHet can improve the delivery ratio by 20 %, increase the throughput by 31 %, and reduce the overhead by 69 % compared to the other benchmarks. In addition, ARHet can achieve a better differentiated service performance.

The all-reduce operation is critically important for communication-intensive workloads emerging at the convergence of High-Performance Computing (HPC) and Internet of Things (IoT) applications. However, existing optimization efforts primarily concentrate on offloading the all-reduce onto network switches, known as In-Network Aggregation, which are incompatible with switchless torus topologies. Driven by our systematic analysis, we identified two key factors that impact the performance of the standard ring all-reduce operation: i. The all-reduce computation process frequently interrupts the GPU/CPU’s computation tasks; ii. The GPU/CPU is, in fact, indifferent to intermediate computational results. Based on this insight, we propose DirectReduce, a fully offloading ring all-reduce architecture that is comprised of three components: (i) the GateKeeper module, responsible for evaluating outgoing data to decide its progression-either directing it to the Protocol Engine for packetization or intercepting it for reduction (e.g., sum, maximum); (ii) the DataDirector module, which classifies the incoming data either is for intermediate result reduction or final result storage; and (iii) the ComputeEnhancer module, designed to execute reduction operations directly on the SmartNIC. Extensive simulation results show that DirectReduce can reduce the ring all-reduce latency by up to 1.98X in a ring (1D-torus) topology, 1.97X in a 3D-torus topology, and 1.75X in a 6D-torus topology compared to the standard ring all-reduce.

With the development of Flying Ad-Hoc Networks (FANETs) technology, FANET routing protocols have attracted more researchers’ attention. Moreover, FANET traffic is usually task-related, so it is crucial to design a task-oriented FANET routing protocol. Named Data Networking (NDN) is a content-centric communication architecture. Compared with the traditional end-to-end communication model, NDN is more concerned about the content and has more potential in task-oriented FANET. However, existing NDN schemes lack awareness of node capabilities, consequently leading to blindly forwarding requests. In addition, traditional broadcast-based forwarding strategies in NDN face fierce competition. To address these issues, we propose a Service-Aware Flying Ad-Hoc Network Routing Protocol with NDN, abbreviated as SARP, which expands the concept of traditional NDN content into services. SARP comprises two components: i) A local-scale service availability information announcement mechanism to perceive Unmanned Aerial Vehicles (UAVs) service capabilities. ii) A task-oriented two-phase forwarding strategy to reduce UAVs’ load and transmission overhead. We compared SARP with existing schemes in ndnSIM and deployed it in the real world. Experimental results show that SARP reduces service response time by 60% and traffic consumption per response by 50% and increases the service response ratio by 20%.

Efficient collective communication, particularly all-reduce operations, is of paramount importance for parallel computing applications like large-scale distributed learning and communication-intensive scientific applications. However, existing optimization efforts for collective communication primarily concentrate on offloading the all-reduce onto network switches, known as In-Network Aggregation(INA), demanding substantial system customization. Unfortunately, these INA solutions are generally incompatible with existing HPC applications that utilize MPI for message passing, except for SHARP and HCOLL.Driven by our systematic analysis, we identified two key factors that impact the performance of the naïve ring all-reduce operation: i. The all-reduce computation process frequently interrupts the GPU/CPU’s computation tasks; ii. The GPU/CPU is, in fact, indifferent to intermediate computational results. Based on these insights, this paper proposes DirectReduce, a fully offloading ring all-reduce architecture that is comprised of three components: (i) the GateKeeper module, responsible for evaluating outgoing data to decide its progression-either directing it to the Protocol Engine for packetization or intercepting it for reduction (e.g., sum, maximum); (ii) the DataDirector module, which classifies the incoming data either is for intermediate result reduction or final result storage; and (iii) the ComputeEnhancer module, designed to execute reduction operations directly on the SmartNIC. Extensive simulation results show that DirectReduce can reduce latency by up to 1.98X in a ring topology and 1.75X in a 6D-torus topology compared to the naïve ring all-reduce.

Nowadays, watching online video streaming through mobile devices is becoming increasingly popular. It is essential to provide users with a satisfying video-watching experience. Multipath QUIC(MPQUIC), an emerging multipath transport protocol, is considered to have great potential in Dynamic Adaptive Streaming over HTTP(DASH). However, we found that existing segment-granularity video streaming transmission in the DASH scenario has a large amount of rebuffering events of short duration. These short-lived rebuffering events seriously impact the user's Quality of Experience(QoE). We found that network fluctuation is the root cause. However, network fluctuation cannot be avoided or predicted. An MPQUIC-based frame-granularity transmission mechanism(MQ-FGTM) is proposed to solve this problem. The frame-granularity transmission avoids the re-buffering events caused by the traditional segment-granularity transmission mechanism that each segment must be downloaded completely before playback. To correctly decode video frames on the client side, we build the stream dependency tree on the server side and construct Minimum Scheduling Groups(MSGs) based on streaming decoding characteristics. Then, we propose a frame-aware scheduler to schedule the MSG in decoding order. Experiments show that MQ-FGTM can reduce the rebuffering number by 68.4%, the rebuffering time by 45.75%, and improve the overall QoE by 21.97%.

Consensus protocols are widely used to build distributed fault-tolerance systems. Unfortunately, existing consensus protocols fail to address communication bottlenecks, and the load imbalance limits the system performance. Named Data Networking is a data-centric network architecture, supporting packet aggregation in the network layer. In this paper, we propose PAN, a weak leader consensus protocol for Named Data Networking. The main idea of PAN is to reduce the message processing load in the application layer and aggregate messages directly in the network layer to eliminate the communication bottleneck. PAN adopts a proactive replicating mechanism to solve the performance bottleneck caused by the replication of large commands. We implement a prototype of PAN in C++ and evaluate it on Mini-NDN. The results show that PAN balances the load well between replicas and the communication bottlenecks are alleviated. Moreover, PAN provides benefits in both throughput and latency compared with other consensus protocols over Named Data Networking.

The gap between the peak-hour Internet and the average level is increasing, which inevitably creates a type of temporary cellular weak coverage when there is a surge in data traffic demand, where any cell-edge device will have a low spectrum efficiency (SE). In this article, we propose a novel incentive architecture based on the dynamic radio frequency charging technology to improve the SE and use the Stackelberg game theory to formulate the problem. In such a model, a small base station (SBS) acts as the leader to offer a desired partition of the resource block obtained by a cell-edge device, while some small energy providers (SEPs) and small virtual access points (SVAPs) that are selected from user equipment act as the followers to make their decisions, respectively, to compete for the free part of such a resource block. Following the potential game rules, all the SEPs compete for a specific free resource part allocated by the SBS, and then, all the SVAPs compete for another nonoverlapping part allocated by the SBS on the basis of the results of the SEPs' potential game. Although our incentive architecture formally has three game stages, it is essentially a two-level Stackelberg game, which is analyzed by using a backward induction method. The theoretical analysis proves the convergence of the above-mentioned game models, and the simulation results demonstrate that the proposed incentive architecture can improve the SE for each cell-edge device.

The ultradensification deploying for cellular networks is a direct and effective method for the improvement of network capacity. However, the benefit is achieved at the cost of network infrastructure investment and operating overheads, especially when there is big gap between peak‐hour Internet traffic and average one. Therefore, we put forward the concept of virtual cellular coverage area, where wireless terminals with high‐end configuration are motivated to enhance cellular coverage quality by both providing RF energy compensation and rewarding free traffic access to Internet. This problem is formulated as the Stackelberg game based on three‐party circular decision, where a Macro BS (MBS) acts as the leader to offer a charging power to Energy Transferring Relays (ETRs), and the ETRs and their associating Virtual Access Points (VAPs) act as the followers to make their decisions, respectively. According to the feedback from the followers, the leader may readjust its strategy. The circular decision is repeated until the powers converge. Also, the better response algorithm for each game player is proposed to iteratively achieve the Stackelberg‐Nash Equilibrium (SNE). Theoretical analysis proves the convergence of the proposed game scheme, and simulation results demonstrate its effectiveness.