Best practices for testing Ethernet & network synchronisation at the cell site
05 September 2016
The 4G mobile subscriber base is continuing to expand and mobile data usage is rising at a phenomenal rate (with some estimates claiming that by the end of 2015 it will have reached 3.7Exabytes/month, which is a 74% increase on what it was at the end of 2014).
(Click here to view article in digi-issue)
These dynamics are giving mobile operators greater impetus to carry out deployment of small cells. Figures from the Small Cell Forum predict that by 2019 as many as 7 small cells will be deployed for every macrocell.
Small cells will markedly improve coverage in areas with large numbers of tall buildings situated in close proximity to one another (combatting the effect of what are often referred to as ‘urban canyons’) as well as boosting coverage in under-served rural environments. They will also increase capacity in high-usage areas with heterogeneous network deployments (being used in conjunction with existing macrocell infrastructure). Small cell implementation is, however, certainly not without its challenges. Their greater prevalence places further pressure on ensuring that the network, and the equipment that supports it, has been installed and activated properly.
Failing to fully test small cell deployments can impact heavily on the mobile network. The sheer number of small cell deployments limits the ability of field test teams responsible to troubleshoot each problem. Furthermore, timing synchronisation issues between small cells and macrocells can lead to dropped calls and degraded macrocell performance - which is exactly the opposite of the intention behind small cells deployment. Test teams need access to instrumentation with an extensive set of features that are suited to small cell deployment. These should include Ethernet backhaul, front haul CPRI and OBSAI, plus legacy interface testing.
Small cell backhaul testing
The aim of small cell backhaul testing is to prevent newly activated small cells from interfering with current macrocell operation and also ensure subscribers’ quality-of-experience (QoE) is kept to a high degree (avoiding dropped calls when transitioning from small cell to macrocell coverage and suchlike). In addition, it can reduce operating expenses associated with troubleshooting small cell deployments.
In a typical small cell site configuration, a small cell eNodeB is connected to an antenna, which broadcasts the radio signal. The small cell has an Ethernet connection for the transmitting/receiving of voice/data traffic (as well as control plane information to coordinate with other cells nearby). In addition to the Ethernet connection, a timing input will also be incorporated into the cell - from which timing information will be acquired. The small cell connects both its Ethernet port and the timing interface to a cell site router (CSR).
Figure 1 describes a typical small cell deployment. There are 3 key test points that field engineers need to address to ensure high quality cell performance. As illustrated in Figure 2, these are the Ethernet interface upstream of the CSR, the Ethernet interface downstream of the CSR and the synthesised timing interface between the CSR and the small cell.
Figure 3 shows that these tests can be grouped into basic and advanced Ethernet backhaul workflows. It is recommended to carry out the basic workflow on every eNodeB installation. The advanced workflow should be performed on a sample number of eNodeB installations or when troubleshooting network timing-related problems. The tests in the basic workflow will allow mobile operators to validate backhaul network performance from the eNodeB to the mobile switching centre (MSC) and also to test for the presence and correct configuration of network synchronisation protocols. The advanced workflow adds tests to measure the quality of the network synchronisation protocols as well as the quality of synthesised reference clocks.
Key test points
1. Upstream of the CSR
The Ethernet link connected to the upstream of the CSR transports Ethernet traffic, as well as providing the timing information needed for the CSR to synthesise the timing reference for the eNodeB. Tests for J-Proof Layer 2 control plane transparency, SyncE configuration and IEEE 1588v2 PTP configuration will be necessitated. The advanced workflow at test point 1 includes measuring wander on a synchronous Ethernet signal to confirm the timing reference to the CSR keep within the confines of a specified mask. It also entails checking connectivity to a 1588v2 grandmaster and measuring PDV and iPDV on the PTP packets with greater accuracy, as well as measuring the one-way delay to/from the PTP grandmaster to ensure upstream/downstream delays are symmetric and/or stable over time.
2. Downstream of the CSR
The Ethernet interface downstream of the CSR connects the CSR to the eNodeB so that the transmission of both user data and signalling information can be achieved. The tests undertaken at this interface ensure robust connectivity and acceptable performance are maintained between the eNodeB and the MSC. Tests at this point include RFC 2544 or Y.1564 (for validating end-to-end configuration at either the Ethernet or IP level), RFC 6349 TrueSpeed (for testing end-to-end throughput using TCP traffic to ensure the network provides the expected throughput without placing additional strain on the limited wireless spectrum). These tests will be described in detail shortly.
3. Synthesised timing interface
The eNodeB obtains its timing reference data from the synthesized timing interface. The accuracy and stability of this timing reference directly determines the accuracy of the frequency transmitted on the air interface. Poor accuracy on the air interface can lead to interference with neighbouring cells and lead to dropped calls or poor data throughput. Critical tests to perform here include time offset and wander measurements on 1 PPS signals and wander measurements on E1, T1, 2MHz, 10MHz signals.
Types of test
1. RFC 2544
RFC 2544 testing is used to verify key performance indicators (KPIs) at the Ethernet or IP level for a single service of data traffic. This test standard calls for measurements of throughput, latency and frame loss. It can be used for assessing Layer 2 or Layer 3 connectivity when only a single stream or single class-of-service (CoS) traffic is present. Due to it only supporting a single stream, RFC 2544 is simpler to configure and faster to run than Y.1564.
Y.1564 is a more advanced test methodology for measuring Ethernet or IP KPIs that can be substituted for RFC 2544 when the network supports multiple CoS traffic. These can include multiple Ethernet VLANs or multiple IP DSCP/TOS values. Y.1564 testing allows verification of both bandwidth profile traffic parameters.
J-Proof is a Layer 2 control-plane transparency test. This can be useful in situations when a mobile operator is implementing an Ethernet virtual private line service to backhaul traffic from a cell site to the MSC. It provides pass/fail indications for each L2CP protocol that successfully traverses the network, as well as indications of header error or payload errors for L2CP frames that return with errors.
4. RFC 6349 TrueSpeed
Measuring throughput at the TCP Layer is vital to maintain a high quality backhaul connection, as data traffic generated from mobile devices predominantly relies on TCP for e-mail, web browsing and mobile application data transfer. RFC 6349 TrueSpeed is a test to measure TCP throughput in both the upstream and downstream directions. TCP throughput (measured at Layer 4) can often be far worse than Ethernet or IP throughput (measured at Layer 2 or 3) because packet loss, network congestion, or changing delay can cause TCP retransmissions. Test parameters will include TCP throughput, TCP efficiency and buffer delay percentage.
5. IEEE 1588v2
IEEE 1588v2 configuration testing provides a way to assess connectivity from the cell site to the PTP master clock by emulating a PTP slave device. In addition, the test measures KPIs for the PTP traffic such as PDV, IPDV and round-trip latency. This test can be run both with the PTP packets contained within Ethernet frames (Layer 2 mode) or with the PTP packets found in UDP segments (Layer 4 mode). The main test parameters here will be connectivity to the PTP master, mean path delay and PDV.
Sourcing the necessary instrumentation for this wide array of different tests can be a challenge too - with lead times, upfront investment and ongoing costs all needing to be considered. Through partnership with Viavi Solutions (formerly JDSU), Microlease can provide test engineering teams with a variety of options by which to gain access to cutting-edge test hardware, as well as offering the necessary technical and logistical support.
Figure 1 - Typical small cell site configuration
Figure 2 - Small cell test points
Figure 3 - Basic & advanced mobility workflows
Figure 4 - Basic test point 1 workflow
Figure 5 - Advanced test point 1 workflow
Figure 6 - Basic test point 2 workflow
Figure 7 - Advanced test point 3 workflow
Contact Details and Archive...