Tech-invite  3GPPspecsRELsGlossariesSIP
Info21222324252627282931323334353637384‑5x

full Contents for  TS 22.261  Word version:   17.2.0

Top   Up   Prev   Next
0…   4…   6…   6.4…   6.8…   6.12…   6.22…   6.26…   7…   8…   A   B   C   D…   F…

 

C  Relation of communication service availability and reliabilityWord-p. 66
Communication service availability and reliability are well known terms used not only within 3GPP but also in vertical industries (IEC 61907 [12]). Communication service availability addresses the availability of a communication service, in vertical applications usually in accordance to IEC 61907 [12]. Reliability addresses the availability of the communication network. The relation of both terms is depicted in Figure C-1.
Up
As depicted, reliability covers the communication-related aspects between two nodes (here: end nodes), while communication service availability addresses the communication-related aspects between two communication service interfaces. In other words, the "gap" between both concepts is the communication interface. This might seem to be a small difference, but this difference can lead to situations, where reliability and communication service availability have different values.
Example: traffic gets "stuck"
The related scenario is depicted in Figure C-2.
Up
This scenario addresses unicast communication from application A to B. The packets are handed over from the application to the communication network at the communication service interface A, and the packets are then transmitted to the end node B. In this example, the packets received by end node B are not exposed at the communication service interface B. So, even if all packets that are handed over to end node A are successfully delivered to end node B within the time constraint required by the targeted service, i.e. even if the reliability is 100%, the communication service availability is 0% since no packets arrive at the "end", i.e. the communication service interface B.
Example: less capacity than agreed
The related scenario is depicted in Figure C-3.
Up
This scenario describes unicast communication of evenly interspersed packets from application A to B. The packets are handed over from the application to the communication network at the communication service interface A, and the packets are then transmitted to the end node B. However, only every second packet is actually successfully handed over to end node A and then transmitted to end node B. Thus, only half of the packets arrive at application B. Note though that the reliability of the mobile network is 100%, since all packets transmitted by end node A are delivered to end node B within the time constraint required by the targeted service. However, depending on the agreed QoS, the communication service availability can be of the same value as the reliability or much lower.For instance, if the agreed survival time is larger than the time between three packets at the communication service interface A, communication service availability and reliability have the same value. However, if - due to the loss of packets at the communication service interface A - the effective bandwidth between application A and B is lower than the agreed value, the communication service availability is 0%.
Note that the shortest time interval over which the communication service availability should be calculated is the sum of maximum allowed end-to-end latency and survival time.
Example: heterogeneous network
The related scenario is depicted in Figure C-4.
Up
This scenario is not in scope for this specification, since it pertains to the particular deployment of a mobile network, but we discuss it nonetheless, as this example provides valuable insight for network operators.
This scenario describes unicast communication from application A to B. The packets are handed over from the application to the communication network at the communication service interface A, and the packets are then transmitted to the end node B. In this example, the packets are transmitted over two daisy-chained networks-one mobile network according to 3GPP specifications, and another network, e.g. a network based on IEEE 802.11n. As in the other examples above, communication service availability is measured between the two communication service interfaces, but the reliability is only measured between end node A and the router node. This has implications for, e.g. the maximum communication latency allowed for each network. In case the agreed end-to-end latency between the service interfaces is, for instance, 100 ms, and the 802.11n network has a latency of 30 ms, the maximum allowable latency for packages in the mobile network is 70 ms (NOTE). So, if the latency in the mobile network exceeds 70 ms, the communication service availability is 0%, despite the agreed QoS stipulating a larger end-to-end latency, i.e. 100 ms.
NOTE:
The transit time through the router node is not considered here. It is assumed to be very small and much less than 100 ms.
Up

Up   Top   ToC