Validating SPEED-5G backhaul solutions – Proof of Concept 3
Work at the backhaul segment of SPEED-5G network architecture aims at (a) increasing the available throughput per link and capacity per area, (b) reducing the hop latency, (c) increasing network availability, and (d) balancing backhaul resources. For realising the above goals, a PtMP wireless backhaul system at 28GHz was enhanced with the following solutions developed in SPEED-5G: (i) channel capacity increase per sector, from 56MHz to 112MHz, (ii) aligned sector collocation, deploying two – and theoretically more – Central Stations, (iii) failsafe capabilities in Central and Terminal Stations with automatic frequency scanning implementing 1:1 protection mode, and (iv) automatic Terminal Station entry at provisioning. PoC3 is designed to enable the testing of these solutions.
The Proof of Concept (PoC) setup is depicted in the figure above. The equipment under test is shown as black boxes, consisting of two Central Stations (CS1 and CS2) and a number of Terminal Stations (TS1, TS2 and TS3 in this case). The CSs and TSs are connected through a 28GHz RF Testbed. A Tester is used to inject both downstream (DS) and upstream (US) traffic, collect it at the other side and measure throughput and latency and BER. A Management Server is also deployed to control and monitor the backhaul system.
The setup also includes two Ethernet switches providing necessary connectivity to the above modules.
Three test cases were run over this setup:
Throughput and latency test case
The objective of this test case is to evaluate the throughput and latency improvements of the SPEED-5G backhaul segment. Downstream traffic is generated from the Tester and injected to the CSs through Switch A. At the other end, the traffic is received from the TSs and forwarded back to the Tester through Switch B. Upstream traffic is injected from the Tester following the opposite path. The Tester measures and reports throughput and latency. Table 1 presents the current results in comparison to the performance of the legacy backhaul system.
Table 1: POC 3 throughput and latency results
|DL throughput per sector||2 Gbps||540 Mbps|
|UL throughput per sector||1.6 Gbps||415 Mbps|
|Aggregate throughput per sector||3.6 Gbps||955 Mbps|
|DL latency||0.984 ms||1.9 ms|
|UL latency||1.0 ms||1.4 ms|
Resource balancing test case
The objective of this test case is to evaluate automatic TS entry during provisioning, based on the load of the CSs. The procedure described below is not only a test case but also actually executed at system power-on when installed. Initially, both CSs are switched on and all TSs are switched off. TS1 is switched on and automatically assigned to one CS (e.g. CS1). Traffic is injected from Tester. TS2 is switched on; it is automatically assigned to CS2 (because CS2 is less busy than CS1). Traffic for TS2 (both DS and US) is injected from Tester. The procedure is repeated for more TSs (e.g. TS3 and TS4). Each TS is automatically assigned to the CS with the lowest traffic load, at the time of its powering. The tests have verified correct behaviour of the above procedure, keeping both CSs balanced with respect to traffic load, verifying resource balancing at the backhaul level.
Network availability test case
The objective of this test case is to evaluate the failsafe capabilities at the backhaul segment of SPEED-5G that were developed in order to increase network availability. Traffic is injected and received from the Tester as in test case (a). CS1 is switched off to emulate failure. Traffic passing through the TSs
associated to CS1 is interrupted. The TSs are automatically switched to CS2 and their traffic is resumed. The availability of a PtMP backhaul sector depends on the Mean Time Between Failures (MTBF) and Mean Time To Repair (MTTR) of its elements (CS and TS). Also, the availability of the sector can be increased when a second CS is added. To calculate it, we consider MTBF = 50 years (a value used by Intracom Telecom, SPEED-5G partner and wireless backhaul vendor, for its commercial systems), a variable MTTR of 3 hrs, 12 hrs, 24 hrs and 72 hrs representing different installation scenarios and operator response times upon failure. Table 2 compares the availability of the SPEED-5Genhanced backhaul versus the legacy system (calculation formulae according to project deliverable D3.2, p. 63).
Table 2: SPEED-5G backhaul availability enhancement
Finally, the service disruption time was measured and compared to a legacy system without failsafe capabilities. As Table 3 shows, for the SPEED-5G backhaul system in case of failure, the disruption is 1 min, independently of the operator’s MTTR, while in the legacy system it is equivalent to MTTR.
Table 3: SPEED-5G service disruption time decrease
|MTTR||Service disruption time
|Service disruption time
|3 hrs||1 min||3 hrs|
|12 hrs||1 min||12 hrs|
|24 hrs||1 min||24 hrs|
|72 hrs||1 min||72 hrs|