The concepts of programmable switches and software-defined networking (SDN) give developers flexible and deep control over the behavior of switches. We expect these concepts to dramatically improve the functionality of switches. In this paper, we focus on the concept of Deeply Programmable Networks (DPN), where data planes are programmable, and application switches based on DPN. We then propose a method to improve the performance of a key-value store (KVS) through an application switch. First, we explain the DPN and application switches. The DPN is a network that makes not only control planes but also data planes programmable. An application switch is a switch that implements some functions of network applications, such as database management system (DBMS). Second, we propose a method to improve the performance of Cassandra, one of the most popular key-value based DBMS, by implementing a caching function in a switch in a dedicated network such as a data center. The proposed method is expected to be effective even though it is a simple and traditional way because it is in the data path and the center of the network application. Third, we implement a switch with the caching function, which monitors the accessed data described in packets (Ethernet frames) and dynamically replaces the cached data in the switch, and then show that the proposed caching switch can significantly improve the KVS transaction performance with this implementation. In the case of our evaluation, our method improved the KVS transaction throughput by up to 47%.

In this letter, a path dynamics assessment asynchronous advantage actor-critic scheduling algorithm (PDAA3C) is proposed to solve the MPTCP scheduling problem by using deep reinforcement learning Actor-Critic framework. The algorithm picks out the optimal transmitting path faster by multi-core asynchronous updating and also guarantee the network fairness. Compared with the existing algorithms, the proposed algorithm achieves 8.6% throughput gain over RLDS algorithm, and approaches the theoretic upper bound in the NS3 simulation.

With multiple network interfaces are being widely equipped in modern mobile devices, the Multipath TCP (MPTCP) is increasingly becoming the preferred transport technique since it can uses multiple network interfaces simultaneously to spread the data across multiple network paths for throughput improvement. However, the MPTCP performance can be seriously affected by the use of a poor-performing path in multipath transmission, especially in the presence of network attacks, in which an MPTCP path would abrupt and frequent become underperforming caused by attacks. In this paper, we propose a multi-expert Learning-based MPTCP variant, called MPTCP-meLearning, to enhance MPTCP performance robustness against network attacks. MPTCP-meLearning introduces a new kind of predictor to possibly achieve better quality prediction accuracy for each of multiple paths, by leveraging a group of representative formula-based predictors. MPTCP-meLearning includes a novel mechanism to intelligently manage multiple paths in order to possibly mitigate the out-of-order reception and receive buffer blocking problems. Experimental results demonstrate that MPTCP-meLearning can achieve better transmission performance and quality of service than the baseline MPTCP scheme.

A Multipath TCP (MPTCP) connection uses multiple subflows (i.e., TCP flows), each of which traverses over a wireless link, enabling throughput and resilience enhancements in mobile wireless networks. However, to achieve the benefits, the subflows are necessarily initialized (i.e., must complete TCP handshakes) and sequentially attached to the MPTCP connection. In the standard (MPTCPST), MPTCP initialization raises several problems. First, the TCP handshake of opening subflow is generally associated with a predetermined network. That leads to degraded MPTCP performance when the network does not have the lowest latency among available ones. Second, the first subflow's initialization needs to be successful before the next subflow can commence its attempt to achieve initialization. Therefore, the resilience of multiple paths fails when the first initialization fails. This paper proposes a novel method for MPTCP initialization, namely MPTCPSD (i.e., MPTCP with SYN duplication), which can solve the problems. MPTCPSD duplicates the first SYN and attempts to establish TCP handshakes for all subflows simultaneously, hence inherently improves the loss-resiliency. The subflow that achieves initialization first, is selected as the first subflow, consequently solving the first problem. We have implemented and extensively evaluated MPTCPSD in comparison to MPTCPST. In an emulated network, the evaluation results show that MPTCPSD has better performance that MPTCPST with the scenarios of medium and short flows. Moreover, MPTCPSD outperforms MPTCPST in the case that the opening subflow fails. Moreover, a real network evaluation proves that MPTCPSD efficiently selects the lowest delay network among three ones for the first subflow regardless of the preconfigured default network. Additionally, we propose and implement a security feature for MPTCPSD, that prevents the malicious subflow from being established by a third party.

Packet losses significantly degrade TCP performance in high-latency environments. This is because TCP needs at least one round-trip time (RTT) to recover lost packets. The recovery time will grow longer, especially in high-latency environments. TCP keeps transmission rate low while lost packets are recovered, thereby degrading throughput. To prevent this performance degradation, the number of retransmissions must be kept as low as possible. Therefore, we propose a scheme to apply a technology called “forward error correction” (FEC) to the entire TCP operation in order to improve throughput. Since simply applying FEC might not work effectively, three function, namely, controlling redundancy level and transmission rate, suppressing the return of duplicate ACKs, interleaving redundant packets, were devised. The effectiveness of the proposed scheme was demonstrated by simulation evaluations in high-latency environments.

This paper aims to establish expressions for IEEE 802.11 string-topology multi-hop networks with transmission control protocol (TCP) traffic flow. The relationship between the throughput and transport-layer function in string-topology multi-hop network is investigated. From the investigations, we obtain an analysis policy that the TCP throughput under the TCP functions is obtained by deriving the throughput of the network with simplified into two asymmetric user datagram protocol flows. To express the asymmetry, analytical expressions in medium access control-, network-, and transport layers are obtained based on the airtime expression. The expressions of the network layer and those of transport layer are linked using the “delayed ACK constraint,” which is a new concept for TCP analysis. The analytical predictions agree well with the simulation results, which prove the validity of the obtained analytical expressions and the analysis policy in this paper.

This paper proposes a transmission control protocol (TCP) for long distance high-speed wireless communications, including free-space optical communications (FSOC). Extreme high frequency of wireless communications enables high-speed bit rate, but frequent signal error, including burst error, can be a quite severe problem for ordinary high-speed TCPs. To achieve 10Gbps or higher data transfer throughput on FSOC, the proposed TCP (designated “TCP-FSO”) has improved and new features including multi-layer congestion control, retransmission control with packet loss point estimation, delay-based ACK congestion control, and ACK retransmission control. We evaluated data transfer throughput of TCP-FSO and the other TCPs, by throughput model analysis and experiment on real implementation. Obtained results show that TCP-FSO achieves far higher data transfer throughput than other high-speed TCPs. For example, it achieved a thousand times higher throughput than the other high-speed TCPs in a real FSOC environment.

In mobile and wireless networks, controlling data delivery latency is one of open problems due to the stochastic nature of wireless channels, which are inherently unreliable. This paper explores how the current best-effort throughput-oriented wireless services might evolve into latency-sensitive enablers of new mobile applications such as remote three-dimensional (3D) graphical rendering for interactive virtual/augmented-reality overlay. Assuming that the signal propagation delay and achievable throughput meet the standard latency requirements of the user application, we examine the idea of trading excess/federated bandwidth for the elimination of non-negligible delay of data re-ordering, caused by temporal transmission failures and buffer overflows. The general system design is based on (i) spatially diverse data delivery over multiple paths with uncorrelated outage likelihoods; and (ii) forward packet-loss protection (FPP), creating encoding redundancy for proactive recovery of intolerably delayed data without end-to-end retransmissions. Analysis and evaluation are based on traces of real life traffic, which is measured in live carrier-grade long term evolution (LTE) networks and campus WiFi networks, due to no such system/environment yet to verify the importance of spatial diversity and encoding redundancy. Analysis and evaluation reveal the seriousness of the latency problem and that the proposed FPP with spatial diversity and encoding redundancy can minimize the delay of re-ordering. Moreover, a novel FPP effectiveness coefficient is proposed to explicitly represent the effectiveness of EPP implementation.

Cloud computing, which enables users to enjoy various Internet services provided by data centers (DCs) at anytime and anywhere, has attracted much attention. In cloud computing, however, service quality degrades with user distance from the DC, which is unfair. In this study, we propose a bandwidth allocation scheme based on collectable information to improve fairness and link utilization in DC networks. We have confirmed the effectiveness of this approach through simulation evaluations.

The growth of Internet traffic and the variety of traffic classes make network performance extremely difficult to evaluate. Even though most current methods rely on complex or costly hardware, recent research on bandwidth sharing has suggested the possibility of defining evaluation methods that simply require basic statistics on aggregated link utilization, such as mean and variance. This would greatly simplify monitoring systems as these statistics are easily calculable from Simple Network Management Protocol (SNMP) calls. However, existing methods require knowledge of certain fixed information about the network being monitored (e.g. link capacities). This is usually unavailable when the operator's view is limited to its share of leased links or when shared links carry traffic with different priorities. In this paper, departing from the analysis of aggregated link utilization statistics obtainable from SNMP requests, we propose a method that detects traffic degradation based on link utilization samples. It does not require knowledge of the capacity of the aggregated link or any other network parameters, giving network operators the possibility to control network performance in a more reliable and cost-effective way.

The Transmission Control Protocol (TCP) with Network Coding (TCP/NC) was proposed to introduce packet loss recovery ability at the sink without TCP retransmission, which is realized by proactively sending redundant combination packets encoded at the source. Although TCP/NC is expected to mitigate the goodput degradation of TCP over lossy networks, the original TCP/NC does not work well in burst loss and time-varying channels. No apparent scheme was provided to decide and change the network coding-related parameters (NC parameters) to suit the diverse and changeable loss conditions. In this paper, a solution to support TCP/NC in adapting to mentioned conditions is proposed, called TCP/NC with Loss Rate and Loss Burstiness Estimation (TCP/NCwLRLBE). Both the packet loss rate and burstiness are estimated by observing transmitted packets to adapt to burst loss channels. Appropriate NC parameters are calculated from the estimated probability of successful recoverable transmission based on a mathematical model of packet losses. Moreover, a new mechanism for coding window handling is developed to update NC parameters in the coding system promptly. The proposed scheme is implemented and validated in Network Simulator 3 with two different types of burst loss model. The results suggest the potential of TCP/NCwLRLBE to mitigate the TCP goodput degradation in both the random loss and burst loss channels with the time-varying conditions.

Streaming services that use TCP have increased; however, throughput is unstable due to congestion control caused by packet loss when TCP is used. Thus, TCP control to secure a required transmission rate for streaming communication using Forward Error Correction (FEC) technology (TCP-AFEC) has been proposed. TCP-AFEC can control the appropriate transmission rate according to network conditions using a combination of TCP congestion control and FEC. However, TCP-AFEC was not developed for wireless Local Area Network (LAN) environments; thus, it requires a certain time to set the appropriate redundancy and cannot obtain the required throughput. In this paper, we demonstrate the drawbacks of TCP-AFEC in wireless LAN environments. Then, we propose a redundancy setting method that can secure the required throughput for FEC, i.e., TCP-TFEC. Finally, we show that TCP-TFEC can secure more stable throughput than TCP-AFEC.

In general, Transmission Control Protocol (TCP), e.g., TCP NewReno, considers all losses to be a sign of congestion. It decreases the sending rate whenever a loss is detected. Integrating the network coding (NC) into protocol stack and making it cooperate with TCP (TCP/NC) would provide the benefit of masking packet losses in lossy networks, e.g., wireless networks. TCP/NC complements the packet loss recovery capability without retransmission at a sink by sending the redundant combination packets which are encoded at the source. However, TCP/NC is less effective under heavy and bursty loss which often occurs in fast fading channel because the retransmission mechanism of the TCP/NC entirely relies on the TCP layer. Our solution is TCP/NC with enhanced retransmission (TCP/NCwER), for which a new retransmission mechanism is developed to retransmit more than one lost packet quickly and efficiently, to allow encoding the retransmitted packets for reducing the repeated losses, and to handle the dependent combination packets for avoiding the decoding failure. We implement and test our proposal in Network Simulator 3. The results show that TCP/NCwER overcomes the deficiencies of the original TCP/NC and improves the TCP goodput under both random loss and burst loss channels.

The transmission control protocol with a random early detection (TCP/RED) is an important algorithm for a TCP congestion control [1]. It has been expressed as a simple second-order discrete-time hybrid dynamical model, and shows unique and typical nonlinear phenomena, e.g., bifurcation phenomena or chaotic attractors [2], [3]. However, detailed behavior is unclear due to discontinuity that describes the switching of transmission phases in TCP/RED, but we have proposed its analysis method in previous study. This letter clarifies bifurcation structures with it.

Computer networks require sophisticated control mechanisms to realize fair resource allocation among users in conjunction with efficient resource usage. To successfully realize fair resource allocation in a network, someone should control the behavior of each user by considering fairness. To provide efficient resource utilization, someone should control the behavior of all users by considering efficiency. To realize both control goals with different granularities at the same time, a hierarchical network control mechanism that combines microscopic control (i.e., fairness control) and macroscopic control (i.e., efficiency control) is required. In previous works, Aida proposed the concept of chaos-based hierarchical network control. Next, as an application of the chaos-based concept, Aida designed a fundamental framework of hierarchical transmission rate control based on the chaos of coupled relaxation oscillators. To clarify the realization of the chaos-based concept, one should specify the chaos-based hierarchical transmission rate control in enough detail to work in an actual network, and confirm that it works as intended. In this study, we implement the chaos-based hierarchical transmission rate control in a popular network simulator, ns-2, and confirm its operation through our experimentation. Results verify that the chaos-based concept can be successfully realized in TCP/IP networks.

Lossy communication networks may be one of the most challenging issues for Transmission Control Protocol (TCP), as random loss could be erroneously interpreted into congestion due to the original mechanism of TCP. Network coding (NC) promises significant improvement in such environment thanks to its ability to mix data across time and flows. Therefore, it has been proposed to combine with TCP called TCP-NC by MIT. In this paper, we dedicated to quantifying the R, a key parameter for redundant packets, and make it close to the loss rate as much as possible, which has not been considered in the previous research. All of these are done by the sender who is completely unconscious of the network situation. Simulation results by NS2 under both wired and wireless networks showed that our method retains all the advantages of TCP-NC, and meanwhile outperforms TCP-NC and the other TCP variants in time-varying lossy networks.

In this paper, we propose a new cross-layer scheme Cooperation between channel Access control and TCP Rate Adaptation (CATRA) aiming to manage TCP flow contention in multi-hop ad hoc networks. CATRA scheme collects useful information from MAC and physical layers to estimate channel utilization of the station. Based on this information, we adjust Contention Window (CW) size to control the contention between stations. It can also achieve fair channel access for fair channel access of each station and the efficient spatial channel usage. Moreover, the fair value of bandwidth allocation for each flow is calculated and sent to the Transport layer. Then, we adjust the sending rate of TCP flow to solve the contention between flows and the throughput of each flow becomes fairer. The performance of CATRA is examined on various multi-hop network topologies by using Network Simulator (NS-2).

In typical end-to-end recovery protocols, an ACK segment is delivered to a source node over a single path. ACK loss requires the source to retransmit the corresponding data packet. However, in underwater wireless sensor networks which prefer flooding-based routing protocols, the source node has redundant chances to receive the ACK segment since multiple copies of the ACK segment can arrive at the source node along multiple paths. Since existing RTO calculation algorithms do not consider inherent features of underlying routing protocols, spurious packet retransmissions are unavoidable. Hence, in this letter, we propose a new ACK loss-aware RTO calculation algorithm, which utilizes statistical ACK arrival times and ACK loss rate, in order to reduce such retransmissions.

High-speed wireless access technologies have evolved over the last years setting new challenges for TCP. That is, to effectively utilize the available network resources and to minimize the effects of wireless channel errors on TCP performance. This paper introduces a new TCP variant, called TCP-BIAD aiming at enhancing TCP performance in broadband wireless access networks. We provide analytical expressions for evaluating the stability, throughput, fairness and friendliness properties of our proposal, and we validate our results by means of computer simulations. Initial results presented in this paper show that this approach achieves high network utilization levels in a wide range of network conditions, while maintaining an adequately fair and friendly behavior with respect to coexisting TCP flows.

We analyze the defects of window-based TCP algorithm in datacenter networks and propose Rate-based Datacenter TCP (RDT) algorithm in this paper. The RDT algorithm combines rate-based congestion control technology with ECN (Explicit Congestion Notification) mechanism of DCTCP. The experiments in NS2 show that RDT has a potential to completely avoid TCP incast collapse in datacenters and inherit the low latency advantages of DCTCP.