Distributed Congestion Control Mechanism Research Articles

Vivace-Distributed: A novel congestion control mechanism for JointCloud environments

As the customer base of cloud computing services continues to grow, the demand for providing cloud resources in an economic manner has become increasingly important. To incorporate multiple clouds, JointCloud has been proposed for sharing cloud services. However, efficient communication among clouds is essential to enable resource sharing in JointCloud, which poses a critical challenge. The existing congestion control mechanisms for JointCloud have been founded to be inadequate in sharing network resources effectively and fairly, resulting in performance shortcomings. The maximum achievable transmission rate is limited to only 50% of the bandwidth, and the minimum and maximum transmission rates for different clouds in JointCloud may have a difference of up to ten times in magnitude.We propose a novel distributed congestion control mechanism for the JointCloud environment, which leverages ideas from rich literature on distributed online learning. The proposed mechanism operates in a distributed manner, with an aggregative variable representing the network condition integrated into each sender’s utility function. Our theoretical analyses establish that the proposed mechanism achieves a unique Nash equilibrium, enabling fair resource allocation among all senders. Furthermore, our simulation evaluations demonstrate that the proposed mechanism outperforms the existing congestion control mechanism in JointCloud environment by achieving a 20% higher throughput in the presence of at least 1% random loss, a 50% higher throughput for short TCP flows, and a throughput ratio of 1 when sharing bandwidth with conventional mechanisms, which indicates a perfect TCP friendliness towards other mechanisms. In summary, our proposed congestion control mechanism is the first to work in a distributed manner, which improves the efficiency and fairness of the congestion control mechanism. The proposed mechanism has the potential to significantly enhance the performance of communication among clouds, making a noteworthy contribution to cloud resource sharing.

A machine learning based Distributed Congestion Control Protocol for multi-hop wireless networks

The application areas of multi-hop wireless networks are expected to experience sustained growth in the next years. This growth will be further supported by the current possibility of providing low-cost communication capabilities to any device. One of the main issues to consider with this type of networks is congestion control, that is, avoiding an excessive volume of data traffic that could lead to a loss of performance. In this work, a distributed congestion control mechanism is proposed for generic multi-hop networks. Different categories of data traffic are taken into account, each of them with different quality of service requirements. The mechanism is based on machine learning techniques, specifically, the CatBoost algorithm that uses gradient boosting on decision trees. The obtained decision trees are used to predict whether the packets to be transmitted over the network will reach their destination on time or not. This prediction will be made based on the network load state, which will be quantified by means of two parameters: the utilization factor of the different transmission channels, and the occupancy of the buffers of the network nodes. To make the values of these parameters available to all nodes in the network, an appropriate dissemination protocol has also been designed. Besides, a method to assign different transmission priorities to each traffic category, based on the estimation of the network resources required at any time, has also been included. The complete system has been implemented and evaluated through simulations, which show the correct functionality and the improvements obtained in terms of packet delivery ratio, network transit time, and traffic differentiation.

A fair and distributed congestion control mechanism for smart grid neighborhood area networks

The need for significant improvements in the management and efficient use of electrical energy has led to the evolution from the traditional electrical infrastructures towards modern Smart Grid networks. Taking into account the critical importance of this type of networks, multiple research groups focus their work on issues related to the generation, transport and consumption of electrical energy. One of the key research points is the data communication network associated with the electricity transport infrastructure, and specifically the network that interconnects the devices in consumers’ homes, the so-called Neighborhood Area Networks (NANs). In this paper, a new fair and distributed congestion control mechanism for NANs is proposed, implemented and evaluated. The main goal of this mechanism is to provide fairness in the access to the network, thus avoiding that some network nodes monopolize the use of the channels due to their higher traffic generation rate, or to their geographical position. Besides, different priorities have been considered for the traffic flows transmitted by different applications. The goal here is to provide the needed Quality of Service (QoS) to all traffic flows, especially when the traffic load is high. The proposal is evaluated in the context of a wireless ad hoc network composed by a set of smart grid meter devices. Applying our proposed congestion control mechanism leads to performance improvements in terms of packet delivery ratio, network throughput fairness between different traffic sources, packet network transit time and QoS provision.

A new distributed congestion control mechanism for networks on chip

Congestion is an important issue in networks and significantly affects network performance. Various congestion control mechanisms have been proposed for the Internet, interconnection networks, etc. However, they are not suitable for network---on-chip systems. Based on the requirements of chip designs, we propose a new distributed congestion control mechanism for network---on-chip systems in this paper. The mechanism uses a new detection metric, the length of the occupied source buffer, to detect network congestion. The new metric is more accurate compared with others. Using the new metric, the congestion information can be directly obtained inside a node. This allows the mechanism to be fully distributed and without requiring any global information. In addition, the presence of real time traffic is considered. Throttling is not required for such traffic to provide QoS. An asymmetric router architecture with additional congestion control unit is also proposed to facilitate the implementation of the new congestion control mechanism. The overhead is relatively low, about 1.79% overhead in area and 2.06 mW in power consumption. The simulations are carried out in OPNET. The results show that our congestion control mechanism alleviates performance degradation for loads beyond saturation, and maintains adequate levels of throughput at higher loads. The new mechanism achieves better network performance than others under different traffic patterns and network sizes.