An autonomic group communication protocol for distributed applications

Current group communication services have mostly been implemented on a homogeneous, distributed computing environment. This limits their applicability because most modern distributed computing environment are heterogeneous in nature. This paper describes the design, implementation, and performance evaluation of a CORBA group communication service. Using CORBA to implement a group communication service enables that group communication service to operate in a heterogeneous, distributed computing environment. To evaluate the effect of CORBA on the performance of a group communication service, this paper provides a detailed comparison of the performance measured from three implementations of an atomic broadcast protocol and a group membership protocol. Two of these implementations use CORBA, while the third uses UDP sockets for interprocess communication. The main conclusion is that heterogeneity can be achieved in group communication services by implementing them using CORBA, but there is a substantial performance cost. This performance cost can be reduced to a certain extent by carefully choosing a design and tuning various protocol parameters such as buffer sizes and timer values.

Group communication services typically generate large multicast data streams. Delivering such massive data streams to the end system nodes at the edge of the Internet has been a challenging problem in terms of high stress on the network links and high demand on network resources and routing node capacities. Most of existing research has been dedicated on geo-distance based routing with various optimizations to alleviate the performance impact on geo-distance based routing due to unpredictable network dynamics. Most representative techniques are targeted at reducing the delivery path length or optimizing routing path by utilizing network locality. In this paper, we identify the inefficiency of geo-distance based routing protocols in many existing multicast overlay networks in terms of both resource utilization and group communication efficiency. To address this issue, we develop a utility-based routing scheme (UDR) that can provide efficient group communication services in a decentralized geographical overlay network. Our approach makes three unique contributions. First, we introduce a utility function to refine the geo-distance based routing in such a way that the routing path selection can carefully incorporate both geo-distance based metric and the network latency. Second, we enhance our utility driven routing scheme with self-adaptive capability by considering the nodes' state and network density. Thus, nodes in the multicast network can dynamically accommodate the changes of network conditions based solely on their local knowledge about the network. Third, we devise a suite of optimization techniques to minimize the maintenance cost and computational complexity of our self-adaptive and utility-drive routing scheme. We evaluate our approach through extensive experiments based on a realistic network topology model and show that the UDR method is highly scalable and it effectively enhances the multicast delivery efficiency for large scale group communication services compared to existing geo-distance based routing protocols.

In distributed systems, high service availability can be achieved by letting a group of servers replicate the service state; if some servers fail, the surviving ones know the service state and can continue to provide the service. Group communication services, such as membership and atomic broadcast, have been proposed to solve the problem of maintaining server state replica consistency. Group membership achieves agreement on the history of server groups that provide the service over time, while atomic broadcast achieves agreement on the history of state updates performed in each group.

In this paper, we propose the design of an extension of the Java distributed object model based on the group communication paradigm for building replicated fault-tolerant servers. Maintaining the correctness and integrity of replicated servers is supported by the group manager object instantiated with each replica to form a logical group and the group membership algorithms to maintain a consistent group view and a reliable multicast mechanism to communicate with other group managers. Group manager object enables the creation of groups of remote objects that cooperate towards some common goal using a join/leave group communication service. Remote object groups simulate the behavior of standard remote objects by implementing a set of remote interfaces and by enabling clients to remotely invoke the methods defined in these interfaces through the standard Java RMI mechanism. For the first time, communication among senders forming a group is based on group method invocations as well, offering a completely object-oriented framework for programmers. And then, it is completed by a state merging service, which simplifies the development of state merging mechanism needed to re-establish a consistent state after the disappearing of join/leave properties

Software based Distributed Shared Memory (DSM) systems have been the focus of considerable research effort, primarily in improving performance and consistency protocols. Unfortunately, computer clusters present a number of challenges for any DSM systems that are not solvable through consistency protocols alone. These challenges relate to the ability of DSM systems to adjust to load fluctuations, computers being added/removed from the cluster, to deal with faults, and the ability to use DSM objects larger than the available physical memory. We present here a proposal for the Synergy Distributed Shared Memory System and its integration with the virtual memory, group communication and process migration services of the Genesis Cluster Operating System.

Mobile agent group concept is a versatile software paradigm capable of providing group task execution flexibility and collaborative adaptability in various application fields of distributed networked systems. However, as the size of these fields is rapidly increasing, improving the performance of the agent group communication in Internet-scale infrastructures should be reconsidered to be suitable for their scale. For this purpose, some appropriate inter-agent group communication protocols are required in distributed agent-based systems. Causal order delivery to a broadcast group is a very important issue for these systems. However, the existing message delivery protocols for mobile agent groups may not satisfy this important requirement. This paper presents a new causally ordered group communication protocol for mobile agent groups to consider weaker message ordering consistency than atomic broadcast semantics while ensuring reliability even with the assumption of unreliable networks for exchanging messages to agent groups and agent mobility. The simulation results show the proposed protocol performs better than the existing atomic broadcast protocol in terms of the total message delivery latency.

Cryptographic communication protocols provide confidentiality, integrity and authentication properties for end-to-end communication under strong corruption attacks, including, notably, post-compromise security (PCS). Most protocols are designed for one-to-one communication. Protocols for group communication are less common, less efficient, and tend to provide weaker security guarantees. This is because group communication poses unique challenges, such as coordinated key updates, changes to group membership and complex post-compromise recovery procedures. We need to tackle this complex challenge as a community. Thus, the Internet Engineering Task Force (IETF) has created a working group with the goal of developing a sound standard for a continuous asynchronous key-exchange protocol for dynamic groups that is secure and remains efficient for large group sizes. The current version of the Messaging Layer Security (MLS) security protocol is in a feature freeze, i.e., no changes are made in order to provide a stable basis for cryptographic analysis. The key schedule and TreeKEM design are of particular concern since they are crucial to distribute and combine several keys to achieve PCS. In this work, we study the MLS continuous group key derivation (CGKD) which comprises the MLS key schedule, TreeKEM and their composition, as specified in Draft 11 of the MLS RFC, while abstracting away signatures, message flow and authentication guarantees. We establish the uniqueness and key indistinguishability properties of the MLS CGKD as computational security properties.

A finite-time resilient consensus protocol (RCP) is developed for a connected network of agents, where communication between agents occurs locally, a few of the agents are malicious (MA), and the non-malicious or cooperating (CO) agents do not know the locations of the MA ones. Networks with a single leader and several followers as well as leaderless networks are considered. Agents are modelled with first-order dynamics, and the inputs to each CO agent that enable consensus are designed using the principles of sliding mode control (SMC). An SMC-based consensus protocol (CP), derived using the Laplacian matrix of the graph describing the network that permits consensus amongst CO agents, is first illustrated; that MA agents can prevent consensus with the use of this CP is also discussed. The SMC-based RCP proposed in this paper requires that the CO agents know the bounds, defined by a statistical distribution, of other CO agents, and that the subgraph of the CO agents be connected; knowledge of MA agents is not needed. With these assumptions, in the case of networks with a single MA agent, CO agents can reach a consensus by disregarding information transmitted by this agent if its state violates some statistical properties. On the other hand, multiple MA agents can be disregarded without the need for these assumptions, owing to the relations that hold with the occurrence of sliding mode. The RCP consists of a message-passing algorithm whose basis can be found in the distributed computing literature. The proposed RCP can also be applied for a leader-follower network.

This paper presents several distributed algorithms that cause a token to continually circulate through all the nodes of a mobile ad hoc network. An important application of such algorithms is to ensure total order of message delivery in a group communication service. Some of the proposed algorithms are aware of, and adapt to changes in, the ad hoc network topology. When using a token circulation algorithm, a round is, said to complete when every node has been visited at least once. Criteria for comparing the algorithms include the average time required to complete a round, number of bytes sent per round, and number of nodes visited per round. Comparison between the proposed algorithms is performed using simulation results obtained from a detailed simulation model (with ns-2 simulator).

Object replication is a technique to enhance fault tolerance and high availability. Group communication is a useful mechanism guaranteeing the consistency among replicated objects. We propose a generic group communication framework that allows transparent plug-in of various group communication protocols with no modification of existing CORBA. For this, we extend the Open Communications Interface (OCI) to support interoperability, reusability of existing group communication, and independency on ORB and OS. The proposed scheme is composed of group membership, group IOR, and group multicast. Experimental results show that group object invocation using the proposed scheme performs substantially better than group object invocations using IIOP as the number of members in a group increases.

In hasty growth of communication, security plays a central role in maintaining confidentiality of data in group communication. Keeping the data intended for the group in confidential manner is the most important security feature need to be sustained for the group communication. An efficient group key management mechanism named as Hybrid Broadcast Group Management Protocol (HBGMP) is devised based on the Reverse Function (RF) and Chinese Remainder Theorem (CRT). The distinctive security among the subgroups is ensured by the reverse function and the session ID of each subgroup is calculated by employing Chinese remainder theorem. By contraption, the Session ID using Chinese Remainder Theorem, with which a cohort of n users requires Sub Group Service Provider (SGSP) to do O (n/m) computation for communication and the storage cost are diminished by diverting the computing load of the Group Service Provider (GSP) into the SGSP. The significance of this protocol is the group member needs to store only two different values during the entire life span and also the rekey message is broadcasted which brings down the communication cost to O(1). The protocol is defined generally for any applications in hybrid architecture. The proposed architecture using CRT and Reverse function is scalable for hefty sized dynamically changing group.

The cooperation of autonomous agents is synthesized in order to satisfy demands imposed on a group of agents or on MAS (multi agent systems). Petri nets (PNs), especially place/transition PNs (P/T PNs), are used for modelling the agents. The modular approach is utilized on that way. PN-based methods of discrete event systems (DES) control and supervising are utilized in order to synthesize the cooperation of the agents. The usefulness and applicability of the approach is demonstrated by examples and case studies.KeywordsAgentsbipartite graphscontrolcooperationdiscrete-event systemsPetri netssupervisorsynthesis

This chapter provides details of group communication services as specified in 3GPP Release 12. This includes the architectural framework adopted for unicast and broadcast delivery of voice and data in order to effectively support group communication. It also highlights different use cases and how group communication solutions apply for these use cases.

Service resilience, defined as the continued availability of a service despite failures and other negative changes in its environment, is vital in many systems. It is typically achieved by state machine replication using group communication middleware for coordination of service replicas. In this paper we focus on systems that represent critical data as Web resources identified by Uniform Resource Identifiers (URIs). The best examples of such systems are RESTful web services. We present RESTGroups : a group communication front-end for RESTful web services. Our system is based on Spread --- a popular group communication toolkit. Contrary to Spread and other such toolkits, we represent group communication services as resources on the Web, addressed by URIs. In the paper, we describe the system architecture and the API.

We have developed a cost-effective regional computer communication network system and examined its implementation. The system provides a LAN environment and group communication services to the customers. The main system components are: the intelligent bridge circuit, called the low-end card (LEC), the center brouter (bridging router), and the management server. The LEC is set up in the subscriber's household and provides the popular Ethernet interface (10 Base-T). The leased circuits are connected between the LEC and the center brouter. The LEC and the brouter are controlled by the management server to function as the group communication system. The brouter, which has a multiplex subscriber-port interface, is required to accommodate a lot of customers cost effectively. This paper presents the system design of a large-capacity brouter having a SDH/SONET multiplex-port interface and its operation. The brouter can accommodate up to approximately 2700 subscribers. Ethernet packets are transferred through each subscriber channel between the LEC and the brouter using the HDLC protocol. A group communication mechanism based on packet filtering via the MAC addresses of PCs is presented. The Ethernet packets are filtered by the brouter to eliminate all packets except those from/to the same group members. The group management operation among the center brouter, the LEC, and the management server is discussed.