569985
NEW ZEALAND PATENTS ACT, 1953
No: 569985 Date: 23 July 2008
COMPLETE SPECIFICATION
SIMULCAST RADIO SYSTEM USING TIME SLOTS
We Tait Electronics Limited, a New Zealand company, of 175 Roydvale Avenue, Christchurch 8053, New Zealand, do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:
569985
SIMULCAST RADIO SYSTEM USING TIME SLOTS BACKGROUND TO THE INVENTION
This invention relates to a method of operating a simulcast radio system, in which adjacent base sites broadcast signals using different time slots to avoid interference between the signals when received at a mobile terminal.
Simulcast is a mode of operating a radio system which offers substantially improved spectral
re-use when compared to standard frequency re-use methods. In particular, simulcast uses the same downlink frequency (and optionally uplink frequency) to broadcast signals at all physical base sites.
When the signals arrive at a user terminal, they combine constructively and destructively
depending upon the instantaneous RF environment. This mixing process combined with the capability of the terminal receiver algorithms places an upper limit on the site separation of the base sites. A typical example is C4FM modulation which limits sites separation to a typical figure of approximately 10km.
The limit of site separation comes about as a result of delay spread. Delay spread is a special case of multi-path spread which describes the combining of signals that have travelled over many paths to the receiving antenna. The multi-path spread observed by a terminal is a function of distance and power from each respective antenna.
The limit on site separation means that more sites are required to cover the same geographical area. This introduces substantial cost. Various methods to improve the site separation have been used in the past. These include; alternative modulations, launch time offsets from transmitters relative to one another, control of transmitter powers, directional antennas and improved receiver algorithms.
Some digital protocols such as DMR (ETSI TS 102 361-1) are based on TDMA (time division multiple access). A frame is defined and divided into time slots, each slot being used as a different channel, rather than parts of the same channel. Slots may be 30 ms units for example.
569985
SUMMARY OF THE INVENTION
It is an object of the invention to provide an improved simulcast radio system, or at least to provide an alternative to existing systems.
In one aspect the invention resides in a method of operating a simulcast radio system, including: allocating a radio channel having a channel frequency for broadcasting of signals between base stations and terminals, defining a frame having two or more time slots for use on the channel, setting up a call between two or more terminals in the system, transmitting data from a first base site to one or more terminals in the call, using the channel frequency and a first time slot in the frame, and transmitting the same data to one or more terminals in the call from a second base site adjacent the first site using the channel frequency and a second time slot in the frame.
Preferably the method further includes transmitting the data to the terminal from a third base site adjacent the first and second sites, using the channel frequency and a third slot in the frame.
In a hybrid system the method further includes allocating one or more further radio channels having respective channel frequencies for transmission of signals between base stations and terminals.
In one geographical arrangement the method further includes transmitting the same data to or more terminals in the call from a chain of adjacent base sites using the first and second slots alternately along the chain.
In another aspect the invention resides in a simulcast radio system having a plurality of base sites and a radio channel for broadcasting of signals from the base sites to user terminals, wherein the channel includes a common frequency and frame, and calls are set up between terminals in which adjacent base sites transmit signals containing the same data to the terminals using different slots in the frame.
Preferably the terminals receive signals over the channel from adjacent base sites and process data in different slots as data from different base sites.
569985
In one embodiment the base sites are located in a chain with adjacent sites in the chain using different slots. The base sites are located in a chain with alternate sites in the chain using common slots.
In a further aspect the invention resides in a method of processing signals at a terminal in a simulcast radio system, including: setting up a call between the terminal and one or more other terminals in the system, receiving first and second broadcast signals during the call, from adjacent base sites using a common frequency and common frame, allocating data in a first slot of the frame to the first signal, and allocating the same data in a second slot of the frame to the second signal.
Preferably the method further includes selecting data from one or other of the slots based on signal quality. Alternatively the method involves a diversity approach by combining data from both slots to determine an enhanced signal.
In another aspect the invention resides in a terminal for a simulcast radio system, including a control unit which is programmed to participate in calls with other terminals in the system by processing broadcast signals received from base sites with a common channel frequency and common frame, and to allocate the same data from different slots in the frame to different base sites.
Preferably the control unit compares data from different slots to determine which signal has best quality. Alternatively the control unit combines data from different slots to determine a signal having enhanced quality.
LIST OF FIGURES
Preferred embodiments of the invention will be described with respect to the drawings, of which:
Figure 1 schematically shows a simulcast radio network,
Figure 2 shows coverage for a network in which a physical channel is divided into three time slots,
569985
Figure 3 shows signal transmissions from neighbouring base sites with the channel divided into two slots,
Figure 4a shows timing of the signals in adjacent slots that are synchronised,
Figure 4b shows timing of the signals in adjacent slots that are not synchronised,
Figure 4c shows timing of uplink transmissions,
Figure 5 shows a typical base site transmitter/receiver,
Figure 6 outlines operation of the base site in a two slot configuration,
Figure 7 shows a typical terminal unit,
Figure 8 outlines operation of the terminal unit in a two slot configuration,
Figure 9 shows signal transmissions in a four slot system,
Figure 10 shows signal transmissions in a hybrid system,
Figure 11 shows transmissions from a chain of base sites,
Figure 12 shows a problem for control channel coverage over a wide area, and
Figure 13 shows a method by which control channel coverage can be extended.
DESCRIPTION OF PREFERRED EMBODIMENTS
Referring to the drawings it will be appreciated that the invention can be implemented in a variety of different simulcast systems. The embodiments described here are given by way of example only.
Figure 1 schematically shows a typical simulcast network having a geographically separated set of base sites, shown as transmitter towers. Located at each tower is a frequency/timing reference. GPS is typically used to lock these references together and in this way, the launch times of each base station can be precisely controlled. Attached to each timing/frequency reference is a base station. Typically one base station is required at each site per physical RF channel. A switch is then typically used to network the system together typically using an IP backbone. When an uplink transmission is received (i.e a transmission from a mobile station) then each site that receives the information sends its data sequence to a voter. The voter selects the best (i.e highest quality) data to broadcast on the downlink. This data packet is now sent back to each site where-upon it is re-transmitted from each site on the downlink, (eg. when setting up a call between two or more terminals in the system).
In general terms, the data that is broadcast on the downlink can be sent in alternative slots. By separating the transmissions in this way, collisions on the downlink can be managed to allow alternative system designs. Alternatively, this method can be used to separate geographically
569985
from TDMA slots so allow the same downlink frequency to be used to cover a much wider geographic area, such as for control channels.
Figure 2 shows an example in which a single physical frequency channel contains three time 5 slots allowing slot re-use geographically such that the system can be deployed across a wide geographical area. This capability of deployment is particularly relevant to PMR systems since PMR systems are typically characterised by large coverage requirements for relatively small user densities. In the general case, the physical channel may be divided into N time slots. In this case, N adjacent sites operating on the same downlink frequency can be assigned 10 individual slot numbers from 1 to N. In this way, no multi-path spread occurs and as such, there is no limit on site separation as a result of normal multi-path limits. Each site is assigned a unique slot number and as a result, no interference occurs between adjacent transmitters. By taking advantage of the slotted structure, it is possible to make substantial improvements to the simulcast scenario detailed in the background.
Consider a two site scenario as shown in Figure 3. In this scenario both slot 1 and slot 2 carry the same information but in alternate slots. During the other slots (greyed), the base station ramps down, transmitting no power. In this way, the terminal observes no multipath problem in the received slot. The terminal simply receives repeated information in both slots and 20 makes a decision as to which block to process. This process enables improved received performance since the two paths are uncorrelated. Alternatively, the terminal could combine the information to improve performance further, by diversity. By establishing simulcast transmissions in this way, the site separation of the base stations are no longer limited to the capability of a particular modulation in multi-path. Thus, in scenarios where standard 25 simulcast operation may not be feasible for geographic reasons, the present invention offers an alternative system design that may be deployed.
Figure 4a illustrates an example transmission from two base sites A and B. The transmissions are offset from one another by one TDMA slot. As a result, the receiving terminal does not 30 suffer limited performance as a result of signal collision. Base station A is transmitting on frequency 1 in slot 1. Base station B is transmitting in slot 2 also on frequency 1. In each base power is ramped up at a time appropriate to the slot boundaries. In the case of the DMR standard, this ramping time is defined in the standard. The modulation start times from each transmission (i.e. base A and B) can be precisely timed with respect to one another. A typical
569985
launch time accuracy might be luS or alternatively 0.5% of symbol time. Notably the transmission from Base B is an integer number of symbols later than Base A. This allows the synchronisation information from base A to remain valid for base B from a receiving terminals perspective. Precise control over launch time also enables a receiving terminal to 5 store the signals from slot 1 and slot 2 respectively to then combine them for a diversity benefit. Clearly, N is a function of the slot period. In the case of DMR the slot time is 30ms and the symbols time (Ts) is 1/4800. Thus N in this case would be 144.
Figure 4b illustrates an alternative, in which the transmission time for Base B can be a non-10 integer number of symbols later than Base A. In this option, the launch time accuracy between base stations does not have to be so precisely controlled. A receiving terminal can therefore not assume that the synchronisation information from Base A is valid for Base B. The terminal then must use the synchronisation information within the current slot to decode the current burst. The synchronisation information in the centre of the slot is generally used to 15 decode data prior to and following that sync burst.
Figure 4c shows an example uplink transmission from the terminal. Although the terminals may receive information on both downlink channels, the terminal can uplink in a number of methods. It can uplink in slot 1 only, as illustrated. Alternatively, it can uplink in slot 2 only. 20 Alternatively it can uplink continuously by using both slot 1 and slot 2.
Figure 5 is a system diagram for a typical base station unit. The system contains the following components, and operates generally as follows. Receiver hardware filters and mixes the RF signal down to signal frequencies that can be operated upon within the digital 25 domain of signal processing. This requires knowledge of the current RF frequency being received and this is supplied for the purpose of mixing from the Frequency Generation system. The Frequency generation system creates oscillations of a rate which allow practical mixing to take place of the wanted RF signal. This mixing typically takes place to bring the signal frequency down to a rate which can be quantized for signal processing operation. The 30 Frequency Generation system is typically implemented as a combination of hardware and signal processing. In an RF duplex system such as a base station, both uplink and downlink frequencies are generated.
569985
A receiver signal processing block then filters, mixes, synchronises and decodes the bit stream to a point where binary decisions can be made to allow the digital signal to be passed to the control unit as bursts of data (or data blocks). The control unit receives the decoded blocks of data and acts upon then according to their content. In the system described here-in, 5 the control unit may compare the decoded bursts from slot 1 and slot 2 to establish that they are repeats of one another. The memory contains the implementation code for the algorithms used to decode the wanted signal and process it according to the required algorithms. The User interface allows the user to select a range of options typically available on a radio of this type such as channel, user ID, signal strength, encryption keys, menu navigation, and so on.
When the base station in Figure 5 needs to make a transmission, the data block is created within the control unit. The burst of data is then passed through the signal processing chain to be processed into a waveform suitable for transmission on the air interface. In a DMR system, the key element is the encoding filter which is described as a Route Raise Cosine 15 filter of alpha value 0.2. This shapes the waveform for transmission before it is mixed back up to RF frequencies. The Frequency generation subsystem is used to supply the relevant mixing components to pass the signal up to an intermediate stage ready for processing by the hardware. The transmitter hardware then mixes the signal up to the required RF frequency and amplifies the signal to transmission to the antenna and the radiation across the air 20 interface.
A Duplexer is typically used at the front end of the transceiver. This allows dual reception and transmission of the base station which is typical of base station operation. In other words received signal operating on carrier frequency 1 is passed through to the receiver line-up. 25 Transmitted signal operating on carrier frequency 2 is passed out from the transmitter line-up to the antenna. A high accuracy frequency reference is used to precisely control the base station transmit frequency. A GPS unit is linked to the frequency reference which is used to create a precise timing reference. This timing reference is used to ensure that transmissions from each base station on the system occur at precisely the same time (or an offset time). 30 The GPS provides a 1 PPS (one pulse per second) timing pulse. This provides a reference time with microseconds accuracy.
Figure 6 shows typical operation of a base station. The slot or slots that a base station will transmit within are defined by configuration of the channel. In normal operation the base
569985
station may transmit continuously i.e. in both slots of a DMR channel. Alternatively the base station may transmit within a single slot, which is also described in the DMR specification. This latter mode is named single frequency repeater mode. Referring to the example of Figure 4, base station A may be configured to transmit in slot 1 only. Base station B, 5 geographically located some distance away may be configured to transmit in slot 2.
Figure 7 is the system diagram for a typical terminal unit. The system contains the following components and operates generally in the following manner. The switching equipment is the apparatus required to enable either transmitted signals to pass out to the antenna and thus onto the air interface. Alternatively, signals received directly from the air interface are passed to the receiver line-up. The receiver hardware filters and mixes the RF signal down to signal frequencies that can be operated upon within the digital domain of signal processing. This requires knowledge of the current RF frequency being received and this is supplied for the purpose of mixing from the Frequency Generation system. The Frequency Generation system creates oscillations of a rate which allow practical mixing to take place of the wanted RF signal. This mixing typically takes place to bring the signal frequency down to a rate which can be quantized for signal processing operation. The Frequency Generation system is typically implemented as a combination of hardware and signal processing. The receiver signal processing then filters, mixes, synchronises and decodes the bit stream to a point where binary decisions can be made to allow the digital signal to be passed to the control unit as bursts of data (or data blocks).
The control unit in Figure 7 receives the decoded blocks of data and acts upon them according to their content. In the system described here-in, the control unit may compare the 25 decoded bursts from slot 1 and slot 2 to establish that they are repeats of one another. The memory contains the implementation code for the algorithms used to decode the wanted signal and process it according to the required algorithms. The User interface allows the user to select a range of options typically available on a radio of this type such as channel, user ID, signal strength, encryption keys, menu navigation, and so on
When the terminal in Figure 7 wishes to make a transmission, the data block is created within the control unit. The burst of data is then passed through the signal processing chain to be processed into a waveform suitable for transmission on the air interface. In a DMR system, the key element is the encoding filter which is described as a Route Raise Cosine filter of
569985
alpha value 0.2. This shapes the waveform for transmission before it is mixed back up to RF frequencies. The Frequency generation subsystem is used to supply the relevant mixing components to pass the signal up to an intermediate stage ready for processing by the hardware. The transmitter hardware then mixes the signal up to the required RF frequency 5 and amplifies the signal to transmission to the antenna and the radiation across the air interface. A switch is typically used at the front end of the transceiver unit to select between transmit and receive paths, forming an RF simplex transceiver. However, RF duplex units are equally applicable.
There are two main methods of deploying a system of the kind described above 1. To enable site separation in a pure simulcast system. 2. To enable a wide area deployment of a single downlink frequency such as a control channel.
In mode 1 above a terminal is configured to expect the same data block in both downlink 15 slots (in a two slot scenario such as DMR). As a result, the terminal can either be configured on a per channel basis to expect information in both slots or it can detect this automatically,
In the process of Figure 8, the terminal receiver can detect that the transmissions are repeats of one another. Slots 1 and 2 are received sequentially and data within the slots is then 20 compared. This comparison can happen in a number of ways. One method of comparison is a direct comparison of the decoded bit stream. This will identify if the slots are the same. Alternatively, specific signalling information could be entered into each bit stream to identify a repeat is present. Alternatively, a correlation can be performed between each data set, where a strong correlation indicates that the message is the same. If the slots are a repeat of 25 one another then either the best data block is selected and used or the two received signals are combined. If the data packets are not repeats of one another then each data block is processed in the usual manner of the radio operation. There are a number of ways to combine the received signals that are described in the open publications. These include Equal Gain Combining and Maximum Ratio combining. Once either repeated or independent slot 30 information is established, the data block (or blocks) is processed. Typically one of the first steps in processing is to identify if the data block is addressed to this unit. A simple example of this processing is shown in Figure 8. It is important to note however the actual contents of the data block are independent of the method shown in this apparatus.
569985
In a simpler form, the terminal knows from its channel configuration that the downlink information that the same information will be sent in two consecutive slots. As a result, it receives the first slot of data then the second slot of repeated data. The best quality information is then selected to be processed. The quality of the information is assessed using 5 a number of mechanisms as known in the industry. These mechanisms typically include signal strength, Bit error rate estimate, synchronisation strength etc. By selecting the best signal the terminal receiver is afforded performance benefit.
All transmissions within a simulcast group are managed precisely in time. The transmissions 10 from two base stations geographically separated by for example 10km, will launch their transmissions (notably modulation) within 1 or two microseconds of one another. The terminal is capable of decoding signal arriving with up to a specified delay if the signal powers observed are the same. Alternatively, if the signal powers are larger than OdB then greater delay values can be decoded.
Figure 9 shows an example of a simulcast system in which N=4. Each site is assigned a unique slot number and as a result, no interference occurs between adjacent transmitters separated by distances dl, d2, d3, d4.
Figure 10 shows a simple hybrid system with two simulcast pairs. The site separations within each pair are dl, d2 and have no limit. The slot 1 and slot 2 transmitters are located at distances m3, m4 determined by the modulation limits.
Figure 11 gives a further example in which a geographic chain of simulcast sites is arranged 25 for a railway, bus route, waterway, coastline or similar feature, in which the site separations may be too large for standard simulcast operation. In this case, base 1 and base 3 transmit in slot 1. Base 2 and 4 deploy slot 2. Distances dl, d2 and d3 are set so that the simultaneous transmissions from slot 1 or 2 respectively do not interfere with one another to an extent that causes serious degradation. A terminal close to base 1 will receive a dominant signal from 30 base 1 as opposed to a very weak signal from base 3 for example, known as the capture effect.
In a further example, the method can be applied to control channels of a trunked TDMA system. In a two-slot system, a control channel comprising slot 1 and slot 2 could be
569985
configured such that slot 1 is used by one site and slot 2 is used by a separate site. As a result, the geographical area covered by the single frequency control channel becomes substantially larger. A control channel can therefore be distributed across two physical sites. In this case, the same control channel information is sent in both slots. This allows a mobile station over a 5 wide geographic area to operate on the same control channel.
Figure 12a illustrates an example coverage problem of a control channel. In this example, a single site is shown operating on frequency fl in continuous (non TDMA) mode. This site is transmitting a string of control channel messages (M1,M2,M3...) intended for a population 10 of terminals A. These terminals are receiving on fl. The required coverage area is shown in gray. This required coverage area clearly extends beyond the coverage of the site 1.
One known solution is to insert another site operating on frequency f2 and instruct the terminal to hunt (or scan) for the best control channel. It is now usually difficult to get extra 15 frequency allocations of this kind. As a result we must seek alternative means of covering the wide area with a single frequency.
Figure 12b shows an alternative configuration in which site 2 is also transmitting on frequency fl but its transmissions are interleaved with the transmission of site 1. The 20 messages Ml, M2 M3 etc are transmitted from both sites in adjacent slots. Thus we have coverage over a wide area but using the same frequency. This form of configuration is ideally suited to applications where there is a relatively low user density over substantial coverage areas such as rural installations.
569985