Abstract: When a proxy device is used between a transmission terminal and a reception terminal, receives final data from the transmission terminal, and fails immediately thereafter, a situation is avoided where transmission is completed in the transmission terminal, but reception is not completed in the reception terminal. There is provided a means for not returning ACK when a data packet is received in a state in which there is no data which waits for ACK and transmission of which is in progress, and there is no data which waits for arrangement and reception of which is in progress in the proxy device, for returning ACK for data received one step before when a data packet is received in a state in which there is data which waits for ACK and transmission of which is in progress, or there is data which waits for arrangement and reception of which is in progress.

Abstract: A failure detection protocol is applied so that failure detection requests for the same aggregated link can be processed even when the failure detection requests differ among routing protocols. The failure detection protocol is executed for each of the physical links that form an aggregated link, and a notification is sent to a higher-level application (for example, a routing protocol), which is a requesting source of a communication line failure monitoring request, with a discrimination between a case in which the communication of the aggregated link in its entirety fails and a case in which some physical links of the aggregated link fail.

Abstract: Erroneous detection of a failure in spite of the absence of failure on the route on account of a communication delay is to be prevented. A node device detects a failure on the route to a remote node device is detected by a failure to receive a packet from the remote node device after the elapse of a determined detection time. The node device sets “remote system transmission interval×detection multiplier” to be a first detecting time on the basis of a packet transmission interval determined by negotiation with the remote node device. It also sets the sum of “remote system actual transmission interval ×detection multiplier” and “tolerable delay time” to be a second detecting time. The first detecting time and the second detecting time are compared, and what is longer is determined as the detecting time to be applied to actual operation.

Abstract: A failure detection protocol is applied so that failure detection requests for the same aggregated link can be processed even when the failure detection requests differ among routing protocols. The failure detection protocol is executed for each of the physical links that form an aggregated link, and a notification is sent to a higher-level application (for example, a routing protocol), which is a requesting source of a communication line failure monitoring request, with a discrimination between a case in which the communication of the aggregated link in its entirety fails and a case in which some physical links of the aggregated link fail.

Abstract: Erroneous detection of a failure in spite of the absence of failure on the route on account of a communication delay is to be prevented. A node device detects a failure on the route to a remote node device is detected by a failure to receive a packet from the remote node device after the elapse of a determined detection time. The node device sets “remote system transmission interval x detection multiplier” to be a first detecting time on the basis of a packet transmission interval determined by negotiation with the remote node device. It also sets the sum of “remote system actual transmission interval x detection multiplier” and “tolerable delay time” to be a second detecting time. The first detecting time and the second detecting time are compared, and what is longer is determined as the detecting time to be applied to actual operation.