Takayuki Tachikawa

Group communication protocol for real-time applications

In distributed applications, a group of multiple processes cooperate by exchanging messages. It is critical to support the group of application processes with enough quality of service (QoS) including the ordered delivery of messages. The delay time and the message loss ratio are significant QoS parameters. In Internet applications, the delay time and the loss ratio are significantly different in different communication channels. The authors define a novel causality named /spl Delta/*-causality among the messages to hold in the world-wide environment. They discuss how to transmit messages to the destination processes and how to resolve message loss and delay supporting the /spl Delta/*-causality given the requirements of delay time and message loss ratio.

M-ary Commitment Protocol with Partially Ordered Domain

Distributed applications are realized by the cooperation of multiple processes. A group of the processes have to make consensus to do the cooperation. The processes exchange the values with the other processes to make consensus. The processes are referred to as consent if each process takes one value which satisfies a consensus condition. A dominant relation among the values is defined to show what values the processes can take after taking a value in the consensus protocol. Each process decides what value to be taken after taking one value by using the dominant relation. In this paper, we discuss how to make consensus in a group of multiple processes by using the dominant relation.

A purpose-oriented access control model

In distributed applications, multiple objects cooperate to achieve some objectives. Each object is manipulated through an operation supported by the object and then the operation may further invoke operations of other objects, i.e. nested operations. Purpose-oriented access rules indicate what operation in each object can invoke operations of other objects. Information flow among the objects occurs if the requests and responses of the operations carry data. Only the purpose-oriented access rules which imply legal information flow are allowed. In this paper, we discuss how to specify the access rules so that the information flow occurring in the nested invocation of the operations is legal.

A purpose-oriented access control model for object-based systems

Distributed applications are modelled in an object-based model like CORBA. Here, the system is a collection of objects. The objects are manipulated only through operations supported by themselves. The purpose-oriented model is proposed where an access rule shows for what each subject s manipulates an object o by an operation t of o so as to keep the information flow legal. The purpose of s to access o by t is modelled to be what operation u of s invokes t to manipulate o. That is, the purpose-oriented access rule is specified in the form (s:u, o:t). In the object-based system, on receipt of a request op from an object o/sub 1/, an object o/sub 2/ computes op and then sends back the response of op to o/sub 1/. Here, if the request and the response carry data, the data in o/sub 1/ and o/sub 2/ is exchanged among o/sub 1/ and o/sub 2/. Furthermore, the operations are nested in the object-based system. Even if each purpose-oriented rule between a pair of objects satisfies the information flow relation, some data in one object may illegally flow to another object through the nested invocation of operations. In this paper, we discuss what information flow is legal in the nested invocations in the purpose-oriented model of the object-based system.

Checkpoint and rollback in asynchronous distributed systems

This paper proposes a novel algorithm for taking checkpoints and rolling back the processes for recovery in asynchronous distributed systems. The algorithm has the following properties: (1) multiple processes can simultaneously initiate the checkpointing; (2) no additional message is transmitted for taking checkpoints; (3) a set of local checkpoints taken by multiple processes denotes a consistent global state; (4) multiple processes can initiate simultaneously the rollback recovery; (5) the minimum number of processes are rolled back; and (6) each process is rolled back asynchronously. The number of messages for rolling back the processes is O(l) where l is the number of channels. Therefore, the system is kept highly available by the algorithm presented.

Information flow in a purpose-oriented access control model

In distributed applications, a group of multiple objects are cooperating to achieve some objectives. An object is modeled as a pair of data structures and operations. Each object is manipulated through an operation supported by the object and then the operation may further invoke operations of other objects, i.e., nested operations. The purpose-oriented access rules indicate which operation on each object can invoke operations of other objects. The information flow among the objects occur if the requests and responses of the operations carry some data. Only the purpose-oriented access rules which imply the legal information flow are allowed. We discuss how to test the access rules if the information flow occurring in the nested invocation of the operations is legal.

Multimedia intra-group communication protocol

In distributed applications, a group of multiple application processes have to send and receive multimedia messages by using the high-speed network. The multimedia message at the application level is decomposed into multiple smaller packets which are transmitted in the communication system. In this paper, we discuss the atomic and ordered delivery of messages at the application level rather than the system level. In some multimedia applications, the application processes do not mind if some packets are lost and in what order packets from different processes are received. The application process specifies the minimum receipt ratio /spl epsiv/ (/spl les/1) showing how many percentages of whole data in each message the destination processes have to be received at least. The communication system delivers the packets to the destinations in the group so as to satisfy the receipt ratio /spl epsiv/. The protocol is based on the fully distributed control scheme, i.e. no master controller.

Δ-Causality and ε-delivery for wide-area group communications

In distributed applications, a group of multiple processes cooperate by exchanging messages. It is critical to support the group of application processes with enough quality of service (QoS) including the ordered delivery of messages. The delay time and the message loss ratio are significant QoS parameters. In Internet applications, the delay time and the loss ratio are significantly different in different communication channels. We define a novel causality named @D^*-causality among the messages to hold in the world-wide environment. We discuss how to transmit messages to the destination processes and how to resolve message loss and delay supporting the @D^*-causality, given the requirements of delay time and message loss ratio.

Research: Significantly ordered delivery of messages in group communication

In distributed applications, a group of multiple objects are cooperated. The objects receive messages and then send back the responses while the objects may exchange data messages. Kinds of group communication protocols support the reliable and ordered delivery of messages at the network level. Only messages to be ordered at the application level, not necessarily all messages, are required, to be casually delivered. In this paper, we would like to define the significant order of messages based on the conflicting relation among the requests. We would like to discuss a protocol which supports the significant delivery of messages.

Distributed protocol for selective intra-group communication

In distributed applications, a group of application processes is established and the processes in the group communicate with one another, i.e. intra-group communication. Here, messages have to be reliably and causally delivered to all the destinations. In addition, the processes send messages to any subset of the group at any time. This paper presents an intra-group communication protocol which provides the group of application processes with the selective and causally ordered (SCO) delivery of messages. The SCO protocol is based on the fully distributed control scheme, i.e. no master controller and uses the high-speed one-to-one network where messages may be lost due to the buffer overrun and congestion.