5.1 Setup for terminals
26.1323GPPRelease 18Speech and video telephony terminal acoustic test specificationTS
The general access to terminals is described in figure 1. The preferred acoustic access to GSM, 3G, LTE, NR and WLAN terminals is the most realistic simulation of the "average" subscriber. This can be made by using HATS (head and torso simulator), with appropriate ear simulation and appropriate mountings of handset terminals to the HATS in a realistic but reproducible way. Hands-free terminals shall use the HATS or free field microphone techniques in a realistic but reproducible way.
HATS is described in ITU-T Recommendation P.58 [15], appropriate ears are described in ITU-T Recommendation P.57 [14] (Type 3.3), proper positioning of handsets in realistic conditions is found in ITU-T Recommendation P.64, and the test setups for various types of hands-free terminals can be found in ITU-T Recommendation P.581.
Unless stated otherwise, if a volume control is provided, the setting is chosen such that the nominal RLR is met as close as possible.
The preferred way of testing is the connection of a terminal to the system simulator with exact defined settings and access points. The test sequences are fed in either electrically using a reference codec, using the direct signal processing approach, or acoustically using ITU-T specified devices.
The system simulator shall simulate the access network and core network including the speech encoding/decoding specified for the test (e.g. AMR-NB or AMR-WB) but excluding further transcoding beyond linear PCM, see Figure 1.
Unless specified otherwise for the respective test, the radio conditions on the air interface shall have a block error rate of 0% and the jitter in the IP transport for MTSI-based speech shall be ≤ 1 ms.
NOTE 1: For WLAN connections, an RF shielded room may be one way to achieve these requirements on block error rate and jitter. Otherwise, care should be taken with potential sources of radio interference and their impact.
In case of MTSI-based speech, the reference client shall allow to synchronize to the clock of the device under test and include a de-jitter buffer to equalize possible jitter in the signal received from the UE.
When operating with synchronized clock, the de-jitter buffer shall be a static de-jitter buffer and the jitter buffer management shall not compensate for clock skew. The reference client shall not lose or discard packets, shall not trigger retransmission, and shall not use error concealment or time-warping. The initial jitter buffer size (filling level) shall be higher than the maximum expected network jitter and the maximum jitter buffer size shall be at least twice the initial size. During jitter buffer reset, the de-jitter buffer shall be emptied/filled to the initial buffer size. In case of buffer over- or underruns, the reference client shall give a warning and it shall be reported.
NOTE 2: A static de-jitter buffer is a first-in-first-out (FIFO) buffer which at the beginning buffers packets until a given initial buffer size is reached. Due to changing network delays the filling level of the de-jitter buffer can change, but the sum of network delay and jitter buffer delay is constant (as opposed to an adaptive jitter buffer management). The filling level of the de-jitter buffer represents the de-jitter buffer delay.
For measurements with unsynchronized clock e.g. the measurement of clock skew, jitter buffer over- and underruns cannot be avoided due to the unsynchronized clocks. Under the assumption of jitter-free condition the initial jitter buffer size (filling level) shall be chosen such that the maximum clock skew can be compensated without any loss of packets for a given time. For the measurement of clock skews the jitter buffer size should be chosen such that for clock skew of up to 100ppm no loss of packets due to buffer over- or under-run shall occur for a sequence of 160s.
For LTE and NR connections, the system simulator shall be configured for FDD operation, with a default or dedicated bearer (EPC) and QoS flow (5GC) and reference measurement channel scheduling that provides enough resource block allocation for transmitting a full speech packet within a transmission time interval of 1ms. No HARQ re-transmissions shall occur. TDD operation, TTI bundling, connected DRX and other forms of scheduling (e.g. SPS) are for further study.
The test setup has to ensure proper clock synchronization of the test equipment to the UE. Clock skew shall be negligible and packet loss shall not occur during the test.
NOTE 3: Any clock skew may result in improper delay calculation or in wrong positioning of the analysis window.
NOTE: Connection to PSTN should include electrical echo control (EEC).
Figure 1: Interfaces (MRP, ERP/DRP…, Air interface and Point of interconnect) for specification of terminal acoustic characteristics