FI94809B - Radio channel fading simulator and method for simulating fading - Google Patents

Description

1 94809

The present invention relates to a radio channel fading-5 simulator and a method for simulating fading.

Multipath propagation is a phenomenon that radio waves experience as they travel in an environment with reflective objects (e.g., buildings, cars, trees, people, etc.). Some of the objects may be stationary 10 and some moving. Due to multipath propagation, the radio signal entering the receiver is a linear combination of several radio signals with different phases, amplitudes and delays. Delay differences are due to the fact that when reflecting from different objects, radio waves travel different distances in the medium. Because some of the reflective objects and also the transmitter or receiver or both may be in motion, the relative phases, amplitudes, and delays change all the time. The radio waves coming from different paths can be combined at the receiver either to cancel or amplify each other. Since the state of the channel changes all the time, the signal strength at the receiver changes accordingly. The phenomenon is called fading.

A fading simulator is a device that simulates a fading ra-25 diode channel under laboratory conditions. An ideal radio signal is applied to the input of the device in either baseband mode or RF center frequency, and a fading signal is obtained from the output. Fading simulators are used in performance measurements of radio receivers. Systems in which fading simulators are required are e.g. the pan-European digital mobile telephone system GSM, the corresponding digital cellular telephone systems in the USA and Japan, the DCS-1800 cellular telephone system in the United Kingdom, spread spectrum systems and military road communication systems.

, 94809

The key parameters when handling fading simulators are delay spread, delay resolution, and channel update rate. Delay spread means the delay difference between the shortest and longest significant signal component of the impulse response of a radio channel.

For example, in urban areas, the delay spread is typically of the order of 3 ... 5 us. The requirement for a large delay variance combined with accurate delay resolution results in a technically complex implementation of the simulator to allow the simulation of any delay component from zero to maximum delay with the accuracy indicated by the delay resolution. The number of delay components required in the simulation is the delay variance divided by the delay resolution. Each delay component is multiplied by a complex time-dependent random number. Due to the large number of delay components, a large number of multiplication operations have to be performed.

The trend in telecommunications technology is towards ever higher transmission speeds and bandwidths. In the same direction, spread spectrum technology is also developing. The typical achievable delay resolution with today's CMOS digital technology is of the order of 30 ns. The maximum delays required in outdoor simulations are of the order of 10 ... 20 us and sometimes more. This means about 300 ... 600 delay com-25 points. Currently existing IC circuits can only implement a class of dozens of delay components. The solutions currently in use provide a large delay variance and accurate delay resolution, but at the same time it is not possible to implement any delay component between zero and maximum delay in steps indicated by the delay resolution, but in practice only a part of all possible delays can be implemented.

For example, in simulators of GSM and DCS-1800 systems, the delay resolution is 0.1 us and the maximum delay is 200 us.

35 The number of possible delay components is only 12. A German fading simulator with a delay resolution of 50 ns and a maximum delay of 100 μs is also known. The number of possible delay components is also only 16.

The object of the invention is to provide a fading simulator of a radio channel 5 which has superior features over the prior art. The radio channel fading simulator according to the invention is characterized in that at least two FIR blocks formed by series-delayed blocks are arranged in the FIR filter so that the delayed signal samples are routed through the series-connected delay blocks of each FIR block to the next FIR block. the outputs are interconnected to provide the desired delay and resolution at the output of the FIR filter.

Other preferred embodiments of the invention and the method according to the invention for simulating fading are characterized by what is stated in the claims below.

The invention will now be described in more detail by way of example with reference to the accompanying drawings, in which Figure 1 shows a typical impulse response structure of a radio channel, Figure 2 shows a block diagram of a radio channel fading simulator * 25 in a general case, Figure 3 shows an FIR block according to the invention in Figure 2 fading.

Very often the impulse response of a radio channel is similar in structure to Figure 1, where the response is located. 30 to the coordinate system, where the x-axis represents the delay d and the y-axis the relative amplitude A. The response has several groups of delay components a, b and c, each of which has its own fine structure. Simulation of such a channel structure requires at the same time a very good delay resolution T, the possibility of long delays and the possibility to adjust the inter-group delays Tx. With the fading simulators according to the prior art, the simulation of such a channel structure is hardly possible.

Mathematically, the multipath propagation phenomenon can be illustrated by formula (1).

OF

s (t) - E (A ^ t ^ uit-Ti)) (1) i = l 10 where s (t) - fading signal, simulator output signal u (t) fading simulator input signal Ai (t) = coefficient of delay component i, of time dependent complex number 15 Ti = delay of delay component i N = number of delay components in the model.

It can be seen from formula (1) that the model used is of a so-called FIR filter with time-varying pin coefficients. A FIR filter (Finite Impulse Response fil-20 ter) is a filter in which an impulse response is generated by delaying the signal in K pieces of successive delay elements by multiplying the output signal of each delay element by a factor and adding the inputs thus obtained. The FIR filter K has a finite integer. The impulse response is determined by the magnitude of the vii-25 factors and the above coefficients. The signal and the coefficients are generally complex numbers.

The structure of a conventional broadband fading simulator at the block diagram level is shown in Figure 2. The ideal modulated RF carrier u (t) is mixed in the RF sub-30 in the coefficient 1 for the intermediate frequency IF. In the quadrature sub-mixer 2, the intermediate frequency signal is converted into a complex baseband signal I and Q, which is converted into sample sequences of real and imaginary parts in the A / D converter block 3. The sample sequence is passed down the data bus to FIR block 4. ,, Ί „N in delay block 5a, 5b, 5c, etc. The delayed samples are multiplied by 5 94809 complex coefficients A ± (t) in multiplier blocks 6a, 6b, 6c, etc. The samples are summed in summation block 7. They are still complex numbers. The real and imaginary portions of the samples are applied as the output signal of the FIR block 4 to their own D / A-5 converters in the D / A converter block 8. The analog I and Q signals are mixed to the intermediate frequency in a quadrature upconverter 9. in the mixer block 10. The output signal of the block is a fading signal at the RF center frequency (sig-10 signal s (t) in formula (1)). The controller 11 controls the static operating parameters of the equipment and supplies the coefficients to the multipliers 6. The coefficients are stored, for example, in the memory 12.

The structure of the FIR block of the improved fading simulator according to the invention is shown in Figure 3. With the exception of the FIR-15 block shown, the structure of the fading simulator according to the invention corresponds to the block diagram of Figure 2, so the general operation of the device described above still applies.

The FIR block 22 shown in Figure 3 has blocks 13a-13N, 14a-14N and 15 which functionally correspond to the FIR block 4 of Figure 2. The complex sample sequence to block 13a corresponds to the sample sequence to the FIR block 4 of Figure 2. If the FIR block 13-15 the delay resolution is T (cf. Fig. 2), its longest delay is N * T, where N is the number of delay elements.

After passing through the first FIR block 22, • 25 samples are applied to the delay element 16. Its delay Tx (cf.

Fig. 1) can be set, for example, by means of a command from the controller unit of the equipment. After the delay element 16, a new FIR block 23 follows, which contains blocks 17a-17M, 18a-18M and 19, which functionally correspond to block. 30 con 22 delay multiplier (coefficients here B ^ t)) and summation ’blocks. The delay resolution of the FIR block 23 is T and its longest delay M * T, respectively. In the general case, M is independent of N. At the end of the FIR delay block chain there is an extension output 21. From there, the samples can be applied to the next FIR block (not drawn). Of these FIR blocks, 94809 can be an arbitrary number in a row.

The filtered sample sequences from the FIR blocks are fed to an adder 20. The complex sample sequence of its output corresponds to the sample sequence of Figure 2 to the D / A converter block 8.

The fading simulator according to the invention shown in Fig. 3 is suitable for simulating the impulse response of Fig. 1, since one FIR group can typically simulate one group of delay components (a, b and c in Fig. 1), and inter-group delays can be used to adjust inter-group delays ( Tx in Figure 1).

It will be apparent to those skilled in the art that the various embodiments of the invention are not limited to the above example, but may vary within the scope of the claims below.

Claims (6)

A fading simulator for a radio channel, wherein the impulse response is formed by delay and multiplication by coefficients (Δt), BA (t)) of samples formed from an RF signal in a temporally variable so-called. FIR filter, may be characterized in that the FIR filter has at least two FIR blocks (22a, 23a, 17a ... 17N) formed in series delay blocks (22, 23) so that the signal samples to be delayed are passed through the respective FIR blocks to the following FIR blocks, and that the filtered outputs from the respective FIR blocks (22, 23) have been summed up for each to achieve the desired total delay and resolution. The fading simulator according to claim 1, characterized in that there is a fixed delay circuit (Tx) between the FIR blocks (22, 23.). The fading simulator according to claim 1 or 2, characterized in that a delay circuit (16) with controllable delay (Tx) is provided between the FIR blocks 20 (22, 23). A method of simulating fading in a radio channel, wherein the impulse response is formed by delay and multiplication with coefficients (Aj (t), Bi (t)) of a 25 signal generated samples in a temporally variable so-called. FIR filters, characterized in that the signal samples to be delayed are passed into the FIR filter through the FIR blocks (22, 23) formed by at least two series-coupled delay blocks (13a ... 13N, 17a ... 17N) respectively. 30 blocks, and that the filtered outputs from the respective FIR blocks (22, 23) are summed together to achieve the desired total delay and resolution. Method according to claim 4, characterized in that the signal samples are delayed between The FIR blocks (22, 23) with a fixed delay (Tx). 10 94809 The method of claim 4, wherein the signal samples are delayed between the FIR blocks (22, 23) with a controllable delay (Tx). *