Method for link failure detection and session transfer to a lively link in the multihoming environment of ID/locator split-based networks

Patent metadata

Field	Value
Publication number	US-9628377-B2
Application number	US-201314770822-A
Country	US
Kind code	B2
Filing date	Feb 27, 2013
Priority date	Feb 27, 2013
Publication date	Apr 18, 2017
Grant date	Apr 18, 2017

Abstract

This invention specifies mechanisms for detection of failure of links directly attached to communication devices (hosts and gateways) or lying in the path between these devices. The invention also specifies mechanisms for recovering communication sessions when the failure occurs. The failure in a link connected to the host or to the gateway is detected by using a link layer trigger by the respective node while the gateway failure or failure in a remote link is detected by a signaling mechanism initiated by the host on the basis of two control timers, probe timer and keepalive timer. The edge router and gateway also assist the host to detect the failure and quickly start the recovery process. They perform packet redirection through a lively link to the host, thus reducing or completely avoiding packet losses while the host performs the recovery mechanism to transfer the session from the failed link to the lively link.

First claim

The invention claimed is:

1. Failure detection and recovery method for a network where both user devices, hosts, and network devices, gateways, are connected to the network through multiple interfaces or links, wherein the hosts are connected to different gateways through different links, wherein a first host, Host A, is communicating through HGW (HIMALIS (Heterogeneity inclusion and mobility adaptation through locator ID separation in new generation network) gateway) 1, one of Host A's gateways, with a second host, Host B, wherein Host B is communicating through HGW4, one of Host B's gateways, with Host A, wherein Host A and Host B also have other gateways, HGW2 and HGW3, respectively, which are not yet used for the communication between Host A and Host B but would be used if a failure of Host A's (or Host B's) link connected to HGW1 (or HGW4) is detected or a failure of HGW1's (or HWG4's) upstream link connecting HGW1 (or HGW4) to Edge Routers is detected or a failure HGW1 or HGW4 is detected, the method comprising steps of:
detecting link failure by Host A;
selecting another lively link of Host A and gateway connected through the lively link, HGW2, by Host A;
selecting Global Locator (GLoc) belonging to the lively link as its new GLoc, GLoc2, by Host A;
registering ID and GLoc, ID-to-GLoc4 mapping, of Host B in HGW2, by sending a peer host registration message from Host A, and the ID-to-GLoc4 mapping of Host B being stored in Table of HGW2;
updating Host B with Host A's ID-to-GLoc2 mapping belonging to the lively link by sending a locator update message from Host A;
updating HGW4 connected to Host B with Host A's ID-to-GLoc2 mapping by sending a peer host registration message from Host B; and
updating HNR (host name registry) records by sending an HNR record update message from Host A to replace GLoc1 belonging to the link failure with GLoc2 belonging to the lively link in Host A's ID-to-GLoc mapping stored in the HNR record;
wherein the method further comprises steps of:
detecting upstream link failure by HGW1; and
sending a GLoc unreachable message from HGW1 to Host A.

2. Failure detection and recovery method for a network in accordance with claim 1 , the method further comprising steps of:
detecting failure in HGW1, HGW4, or a path between HGW1 and HGW4 by having functions for continuously monitoring packet sending and receiving instances in Host A and Host B by using two timers, probe timer and keepalive timer; wherein the probe timer is started and the keepalive timer is stopped after sending a packet and keepalive timer is started and probe timer is stopped after receiving packet; if the keepalive timer hits a specified keepalive timeout value, a series of keepalive packets are sent at the interval of half of the previous keepalive timeout value, and if the probe timer hits a specified probe timeout value, a series of probe packets are sent, each in the interval of half of the previous probe timeout value; and if Host A does not receive any response to the probe packets sent from Host A, Host A starts failure detection and session recovery by:
sending a first peer host registration message containing ID-to-GLoc4 mapping, which is the currently used ID to GLoc mapping of Host B to HGW2;
sending a second peer host registration message containing ID-to-GLoc3 mapping, which is the other ID-to-GLoc mapping of Host B, to HGW1;
determining by Host A, HGW1 is down when a peer host registration response is received only from HGW2; and
determining by Host A, HGW4 or the path leading from HGW1 to HGW4 has a trouble when Host A receives a peer host registration response from both HGW1 and HGW2.

Assignees

• Nat Inst Inf & Comm Tech

Inventors

• Kafle Ved
• Harai Hiroaki

Classifications

• H04W76/025

  Electricity · mapped topic

• H04W88/06

  adapted for operation in multiple networks {or having at least two operational modes}, e.g. multi-mode terminals · CPC title

• H04W60/005

  Multiple registrations, e.g. multihoming · CPC title

• H04W76/028

  Electricity · mapped topic

• H04W24/04 Primary

  Arrangements for maintaining operational condition · CPC title