爱立信mgw软件结构及故障处理.pptVIP

END！ STC, DB and Conf support 1+1 redundancy MESC and CH support n+1 redundancy There can be up to 64 VMGws in one MGw. Each VMGw serves one MGC. VMGw is on all MeSCs. If one MeSC goes down, VMGw is still functional. VMGw has state: locked/unlocked Each MeSC needs to know the status of each VMGw RAA acts as devices resource broker it distributes resource reservations from MESC to GRAs and IMRAs. There is one to one relationship between MESC and RAA and MESC uses only the RAA located in the same GPB. There can be from 1 to 10 active RAAs. The device and resource handling is distributed over one or many subracks in a node. The device and resource handling is distributed over one or many subracks in a node. With the previous information about the software architecture in the M-MGw now it is easier to understand how a call works internally in the M-MGw. The following is an example call flow for call with ATM bearer. 1）MSC Server sends the GCP ADD Request message requesting to choose Context ID, Termination ID and Bearer related information (Local Bearer Address and Binding Reference) 2）SS7 stack decodes the message and checks first the Service Indicator (SI) and next the Signalling Point Code (SPC). Based on the SI (H.248) and the SPC, it associates the GCP message to a specific Mtp3b Access Point (Mtp3bAp). Finally the message together with the Mtp3bAp is sent to STC. 3）STC Receives GCP Add Req and forwards to CH for that VMGw. 4）CH decodes the received GCP message and checks Context ID to determine which MeSC to send it to. Since no Context ID is found in the message, CH selects a MeSC using round robin algorithm 5）MeSC receives decoded GCP message and reserves a context and a termination. First 4 bits of Context is the MeSC ID. MeSC then sends the resource reservation request to RAA which is located in the same board as the MeSC. 6）RAA forwards the resource reservation request to GRA or IMRA in the case of Interactive Message. 7）GRA requests reservation of two