Читать книгу Position, Navigation, and Timing Technologies in the 21st Century - Группа авторов - Страница 134

The advent of 5G networks and indoor small cells for cellular coverage/capacity enhancements has spurred the need for indoor time (phase) and frequency synchronization solution on a wide‐area basis in a scalable manner. Today, time/frequency is distributed directly via a GPS module attached to a small cell or a Grand Master, whose primary source is GPS. The Grand Master distributes time over an Ethernet or fiber‐optic cable using the IEEE‐1588 Precision Time Protocol (PTP) [11]. In indoor environments, a GPS signal is not available, and a complementary solution is required. Table 39.2 indicates the current time and frequency synchronization requirements for various wireless telecommunication networking systems [12]. In the table, CDMA denotes the Code Division Multiple Access network, LTE denotes Long Term Evolution networks, and LTE‐TDD refers to the Time Division Duplex version of LTE; LTE‐FDD refers to the Frequency Division Duplex version of LTE; MBMS refers to Multimedia Broadcast Multicast Service; CoMP refers to Coordinated Multi‐point transmission/reception, and eICIC refers to enhanced Inter‐Cell Interference Coordination. The term ppb in the table indicates clock frequency stability in parts per billion. Note the varying levels of time and frequency sync requirements for different technologies.

Due to the high level of synchronization between the beacons and the frequency stability of the MBS clocks, the MBS system can provide time and frequency in GPS‐challenged environments. In addition, when GPS is unavailable in a given geographical area, MBS can continue to provide an accurate relative phase reference and absolute frequency reference. Given the nature of a timing receiver being at a static location, it is also conceivable that one could perform longer integrations that improve both sensitivity and resilience to any jamming.

Table 39.2 Telecom networks phase/frequency sync requirements

Application	Frequency network/air	Phase	Note
CDMA 2000	16 ppb/50 ppb	±3μs to ±10μs
LTE – TDD	16 ppb/50 ppb	±1.5μs	<3 km cell radius
		±5μs	>3 km cell radius
LTE – MBMS (LTE – TDD and LTE – FDD	16 ppb/50 ppb	±10μs	Inter‐cell time difference
LTE‐A CoMP	16 ppb/50 ppb	±0.5μs to ±1.5μs
LTE‐A eICIC	16 ppb/50 ppb	±1.5μs to ±5μs
e‐911 and Location services		±0.1μs
Small cells	Air: 100–250 ppb	±3μs

The concept of operation of an MBS timing receiver is like a GPS timing receiver. The key difference is that an MBS timing receiver can work in GPS‐challenged or GPS‐denied environments.

Conceptually, the MBS timing receiver determines GPS timing (for example, in the form of a 1 pulse per second (1PPS)) as a part of its receiver time bias determination. The MBS timing receiver can operate in self‐survey mode (where it determines its own position and time) or in pre‐surveyed mode (where it uses a pre‐surveyed position and uses measurements to determine time alone). An MBS‐capable receiver can effectively produce a GPS‐time‐stamped 1PPS and a time‐of‐day (ToD) message. In pre‐surveyed mode of operation, the MBS timing receiver would only need to receive a signal from one MBS beacon to find MBS system time and produce a 1PPS and ToD output. The 1PPS and ToD output can be used to drive a Precision‐Time Protocol (PTP) engine in a Grand Master configuration for use in packet networks.

An illustrative architecture of an MBS timing engine is shown in Figure 39.17. The ToAs from the received MBS beacons is computed by the MBS ranging engine, which are then passed to the MBS timing engine. The MBS timing engine smooths and filters any outliers from the ToAs to generate a 1PPS output that is aligned to GPS time and controls an oscillator loop. A GPS‐grade temperature‐controlled crystal oscillator (TCXO) can be used in the control loop to maintain timing. If any holdover (in the absence of MBS signal) is desired, other oscillators such as oven‐controlled crystal oscillators (OCXOs) and atomic clocks can also be used in place of a TCXO.

Figure 39.17 MBS timing receiver architecture.

Various long‐term tests have been conducted by NextNav as a part of operator trials in the MBS coverage footprints in various deep‐indoor locations where GPS is not available. One such test location was a tall building with a deep‐indoor test location where GPS was not available. A reference GPS receiver (with an antenna on the rooftop) was used to measure the performance of MBS against GPS using a time‐interval counter.

A sample result of the timing performance obtained from the MBS signals in those tests is shown in Figures 39.18 and 39.19. Figure 39.18 shows the MBS time interval error (TIE) as measured over 48 h at a deep‐indoor location in a tall building using a time interval counter. A bias of 138 ns was observed, potentially due to multipath. Figure 39.19 shows the maximum time interval error (MTIE) computed for various time intervals τ using the same data. The MTIE (τ) is the maximum change in TIE over an interval τ and is a metric used in telecommunication networks to measure any sudden jumps in reference time pulses. The MBS MTIE compares favorably to the ITU G8271.1 [13] mask defined for time and phase synchronization of telecommunication networks.

Figure 39.18 Time interval error for MBS timing reception measured over 48 h at a deep‐indoor location in a tall building.

Figure 39.19 Maximum time interval error (MTIE) of MBS timing measured over 48 h at a deep‐indoor location in a tall building.