CC BY
21
TO A QUESTION OF INCREASING THE DATA TRANSMISSION RATE OF PHASE RADIO-TECHNICAL DATA TRANSMISSION SYSTEMS OPERATING UNDER STRONG INTERSYMBOL INTERFERENCE IN LINEAR RADIO PATH
ABSTRACT
Volume of information transmitted by radio engineering data transmission systems exponentially increasing from year to year is a current trend of modern society. This leads to considerable problem of creating new systems capable to operate under conditions of constantly increasing requirements to an effective frequency resource usage.
One of the most effective approaches to solving this problem is the conversion to data transmission in the presence of intersymbol interference produced by the presence of linear selective systems in radio engineering data transmission systems. However, the attractiveness of this approach is associated with an increase of complexity of the receiver itself along with an increase of number of interfering symbols. This ultimately arises the issue not only about the expediency of its implementation, but also about its practical feasibility.
In this paper, the issue of increasing the data transfer rate in radio engineering data transmission systems is discussed, in which the suppression of intersymbol interference in linear selective systems of a linear radio path is not performed, and the correctness of the reconstruction of the transmitted information sequence is achieved due to the appropriate selection of resolving time. It is shown that in some cases for large values of the volume of alphabet of channel symbols, the resolving time can be significantly reduced in the case when bandpass filters are used as linear selective systems.
FOR CITATION: Lerner I. M., Kamalletdinov N. N. To a question of increasing the data transmission rate of phase radio-technical data transmission systems operating under strong intersymbol interference in linear radio path. H&ES Research. 2017. Vol. 9. No. 5. Pp. 92-104.
LERNER
Ilya Mikhaylovich1
KAMALLETDINOV Nail Nadyrovich2
1PhD, Associated Professor of Kazan National Research Technical University n.a. A.N. Tupolev - KAI, Kazan, Russia, aviap@mail.ru
2PhD, Associated Professor of Kazan National Research Technical University n.a. A.N. Tupolev - KAI, Kazan, Russia, kamal585@mail.ru
KEYWORDS: PSK-n-signal; bandpass filters; increasing data transfer rate; frequency detuning; radio engineering data transmission systems.
Volume of information transmitted by radio engineering data transmission systems (RDTS) exponentially increasing from year to year is a current trend of modern society. This leads to considerable problem of creating new RDTS capable to operate under conditions of constantly increasing requirements to an effective frequency resource usage [1-4].
One of the most effective approaches to solving this problem is the conversion to data transmission in the presence of intersymbol interference (ISI) produced by the presence of linear selective systems (LSS) in RDTS. However, the attractiveness of this approach is associated with an increase of complexity of the receiver itself along with an increase of number of interfering symbols. This ultimately arises the issue not only about the expediency of its implementation, but also about its practical feasibility.
This feature is most pronounced among algorithms using the maximum likelihood criterion - optimal algorithms of signal receiving in the presence of ISI [1]. Computational complexity of such algorithms increases exponentially, both with an increase in the number of interfering symbols (which occurs with increasing the transmission rate) and with increasing the volume of the alphabet of channel symbols [5-10].
At the same time, the application of methods with a low computational complexity (zero-forcing filter, linear minimum squared estimation equalizer and decision feedback equalizer) reducing ISI level, which allows to perform the consequent signal processing in the absence of communication channel memory [1], are also not without drawbacks [11]. The common disadvantage of all methods mentioned above is their low noise immunity as compared to algorithms using the maximum likelihood criterion and dependence of their noise immunity on shape of amplitude frequency response (AFR) of communication channel [1, 11]. The last feature is the most pronounced in case when AFR is subjected to considerable variations on magnitude in entire frequency range what is typical for AFR of radio channels [1, 11].
Since all the methods listed above don't provide the complete suppression of ISI and don't allow achieving high noise immunity, then it makes sense to consider the operation features of RDTS, which provide receiving and processing of signals in the presence of ISI without its compensation. In this case the reliable signal receiving ensured by the correct choice of resolution time ofLSS ofRDTS [4, 12-13].
RDTS using multi-position phase-shift keying signals with n discrete states (PSK-n-signals) are the most widely spread nowadays. So it is reasonable to examine the operation features of RDTS of this class specifically for PSK-n-signals.
At the same time it is expedient to estimate the operation features of such RDTS using error probability per channel symbol and induced ISI, depending on channel symbol time duration, since it is the second of the most important characteristics of such RDTS along with capacity which was considered in paper [4].
I. PROBLEM STATEMENT
Since we are primarily concerned with the effect of ISI on error probability, we assume that there is no noise in the communication channel. And the communication channel, for which we estimate error probability per channel symbol when receiving the data of message, consists of a transmitter, LSS, a receiver and a decision device (DD).
In this paper let's confine ourselves to considering the LSS as a population of equitype linear bandpass filters, which satisfy the following conditions: ro0/2AQ0 >15, where 2AQ0— is the resultant bandwidth of LSS, where ro0— is an average frequency of the LSS; and their amplitude-frequency and phase-frequency responses possess even and odd symmetry with respect to. We assume that the receiver is an ideal phase detector and, similarly to DD, doesn't introduce additional ISI to ones formed with LSS.
Each symbol of primary signal that impacts on the input of transmitter takes equiprobably one of n phase values of the signal constellation:
Vk = (k + v sign (||0.5«|| - 0.5« - 0.25)-110.5«|| (1)
Here k = 1,«;|| . | | —rounding operation to the nearest integer; sign (.) — signum function; A<pj( = 2% /n — step between adjacent phase values of the signal constellation; ve [0;0.5] parameter defines the initial phase shift of signal constellation rn =2 %v/n.
' sc
The PSK-n-signal generated by the transmitter from the moment t =0 can be represented as follows
s« (t )=E[1(-(r - 1K )-l(t - r Ts )>
r=1
Xexp((((( + Arn)t + 20r +3y -%)) = £[l(t-(r-1)*,)-
r=1
-1( - r t, )] exP ((((®0 +A®)t + v((r -1)-%) )
(2)
where I e N — is the number of transmitted radio pulses of the PSK-n-signal;
zs— transmission time of each radio pulse of the PSK-n-signal (symbol time duration); r-1
l(t) — Heaviside step function; 9r = £ 28q +9sc;
q=1
ro0+ Aro — the carrier frequency of the PSK-n-signal, where Aro — frequency detuning;
90 — resultant constant phase shift introduced by LSS on frequency ro0+ Am;
2Qr = y((r - 1) t) - y((r - 2) x) — phase jump caused by the transmission of r-thradio pulse, where y((r- 1) x) = 2Qr + S. and y((r - 2) x) = 20M + 9M — are initial phases of r-th and (r-l)-th radio pulses of PSK-n-signal at the output of transmitter. If r = 1 the initial phase is y((r - 2) u) = <p , because we assume that until
the time / = 0LSSisina steady state and transmitter generates the harmonic oscillation exp (/'((o>0+ Aro) x+ ysc - q>0)).
In this case PSK-«-signal at the output of a LSS starting from the time t = 0 according to results of [14] and superposition principle can be represented as follows
Sout (t) = k0 ( ( (©0 + A©)) ()exp (( ((©0 + A©) t - Ф0 )) = : k0 ( ((©0 + A©))exp ( (((©0 + A©) t- ф„ ))(exp ( ((( - 6j ))
+I ^r [exp ( ( +er ))- exp (( ( -er ))),
(r = I 2e„ +er +9s,
(4)
q=1
Vrec (rTs ) = Vklk=k
(6)
where k ' el, n : f (k ', r) = min\wme (rxs )-yk| <
kel.i
a0 ^ 0.5 , what allows to exploit the dynamic operation mode of the considered communication channel in full measure.
To define the operation features of the considered RDTS let's consider the following bandpass filters with identical bandwidth asaLSS:
1. One and two staged resonant filter. According to [14] the settling function of one staged filter is defined as
(3)
where Z (t) —is the complex envelope of the PSK-«-signal atLSS output;
Xr = l(t-(r - l))) (t-(r -l), Ara) ■ B30 (t, Ara) — settling function forLSS [14];
k0 ((ra0 + Ara)) = k((ra0 + Ara)) exp(j'90 ) —the resultant complex transmission coefficient of LSS on frequency ra0+ Ara;
È0 (x, p) = 1 - exp (-x)(cos px - j sin px),
and the settling function of two staged filter is defined as
(7)
On the output of phase detector the informative signal — the slowly varying phase of the PSK-«-signal is defined according to (3) as
Vme () = argZ () = arg[exp (( ((i - 01 ))] +
+£ Xr [exp ( ( + 0r )) - exp ( ( - 0r ))]. (5)
The recovery of the primary signal is made by DD based on the decision rule
a <0.5Am,
pm ~ sr
where apm— is the permissible phase settling error on LSS output. The latter condition is required to ensure the uniqueness of DD solution in the absence of noise. The permissible error is related to measured settling error of slowly varying phase by the following expression apm > \ame (rxs )| = (rxs) - y ((r -1) x^ )| and determines the limiting value for the difference taken by modulus between the specified phase value of PSK-«-signal radio pulse on LSS output and the initial phase value of the same radio pulse acting on its input.
The permissible error a is related to reduced permissible error aQby a0 = a /Aqi^ equation. a0 should be chosen so that
Bo (x, p) = 1 - exp (-x)(cos px - j sin px))l + (1 + jp)x), (8)
where x = Aü0t — non-dimensional time;
AQ0 = AQ0/aA, for one staged resonant filter aÄ= 1 and for two staged — aA = 0.644;
p = Ära / AQ = a, Ara/AQ„ = p„a, — non-dimensional fre-
r A 0 r0 A
quency detuning.
2. Single stage bandpass filter on coupled circuits with different values of a degree of circuit coupling. Settling function for a single stage filter is the following [14]:
Bo (x, p) = 1 - exp (- (1 + jp) x))cos ßx + sin ßx j. (9)
Here a = 1.414 for a single stage filter with degree of circuit coupling ß = 1 and aA = 3.11 for degree of circuit coupling ß = 2.41.
I. PROBLEM SOLVING
Due to the validity of transposition principle for LSS and results of [4] it can be stated that the process of phase settling of PSK-H-signal on LSS output, caused by the accidental essence of the emergence of initial phase values in the transmitted information sequence, is statistic stationary starting from the certain moment of time x^, for any symbols starting from the 5-th. Thus, this allows us to estimate the probability of an error depending on symbol time duration in the presence of ISI for the whole sequence starting from the 5-th symbol.
This problem will be solved by numerical simulation, using expressions (l)-(9) — and the following relation for error probability:
= 1
Perr (AQ0Ts ) = I 1(a pm -|Vme (rTs, m )-v((r - J)Ts , m )|)
(10)
where \\ime (rx, m) — value of slowly varying phase measured on LSS output at the end of the r-th radio pulse for the m-th realization of the transmitted information sequence;
y((r-l) xj5 m) — initial phase value of the r-th radio pulse for the rn-th realization of the transmitted information sequence;
g
g — the quantity of realization of information sequence. First of all the time xs' is initially determined from which the differences between the dependencies of the integral distribution function for measured error for all symbols starting from the 5th do not exceed 5%, when the expression (10) is used.
For the rest first four symbols of the sequence, error probability is defined according to the following relation
Perr (^0 Ts, f )_ Z !(apm (fTs'm )-v((f - !)Ts > m ))
_ 1 — m= 1_
g
where yme (f xs, m) — value of slowly varying phase measured on LSS output at the end of the/-th radio pulse for the m-th realization of the transmitted information sequence;
V ((/- 1) x, m) — initial phase value of the/-th radio pulse for the m-th realization of the transmitted information sequence.
The following parameters were used in the process of numerical simulation: number of transmitted radio pulses I = 30; number of realizations of PSK-n-signal g = 5 x 104; number of discrete states was « = 4,8, 16, 32, 64, 128 with initial shift of the signal constellation 9sc = {0; 0.5%/n; %/n}; the value of reduced permissible error a0 was 0.4999; r=9.
Results of numerical simulation for single- and two staged resonant filter are presented on (fig. 1-9)
Fig.l. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: tuned single-stage resonant filter (1-6); tuned two-stage resonant filter (7-12). Initial phase shift ofPSK-ra-signal q>jc: % In (solid line); 0 {dashed line). Number ofdiscrete states n: 4 (1,7); 8 (2,8); 16 (3,9); 32 (4,10); 64 (5,11); 128 (6,12)!
P„<A*VJ
Û.6
0.4
0.2
1 1 j 6
v \ ' "^-N 4
0.5
1.5
2.5
3.5
Fig.2. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: single-stage resonant filter. Non-dimensional frequency detuning p0 = -0.5. Initial phase shift ofPSK-n-signal q>jc: % In (solidline); 0.5% In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Analyzing dependencies of error probability for resonant filters (fig. 1-9), we can make the following conclusions:
1. The presence of frequency detuning leads to an increase in the resolving time of the resonance filters. At the same time, the degree of increase is more pronounced, the higher the values of frequency detuning.
2. The best signal constellation with an even number of discrete states, ensuring the absence of a constant error during PSK-n-signals receiving and processing in the presence of ISI is that whose m = %ln.
T sc
3. The presence of frequency detuning doesn't lead to change of configuration of resolving time relations between different PSK-H-signals relatively to case of the absence of frequency detuning.
4. Changes in error probability with positive values of frequency detuning for PSK-H-signals considered in this paper,
when the value of initial phase shift of signal constellation differs from q>sc = n/n is more pronounced as compared to the case when it takes negative values.
Results of numerical simulation for single stage bandpass filter on coupled circuits with different values of a degree of circuit coupling are presented on (fig. 10-19).
Analyzing dependencies of error probability for bandpass filter (fig. 10-19), we can make the following conclusions in addition to those that were made above for resonance filters:
1. Unlike resonant filters, the dependence of the error probability of symbol reception caused by ISI contains additional "transparency windows" at which it is zero, which allows to increase the transmission rate of PSK-H-signals. In the absence of frequency detuning the degree of manifestation of this property increases (a decrease in the value of the number of discrete states occurs when this effect is manifested) with an
Fig. 3. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: single-stage resonant filter. Non-dimensional frequency detuning p0 = 0.5. Initial phase shift ofPSK-n-signal : % In (solid line); 0.5% In {dashedline); 0 (dot-dashedline). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (5); 128 (6).
Fig. 4. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: single-stage resonant filter. Non-dimensional frequency detuning p0 = -1. Initial phase shift ofPSK-n-signal. q>jc: % In (solid line); 0.5% In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (5); 128 (6).
____
s. 6
-^v 5
V ■ 1 \ - V
\ , ^ %......-.
\ \ Vv„
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Ml0ts 6
Fig. 5. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: single-stage resonant filter. Non-dimensional frequency detuning p0 =1. Initial phase shift ofPSK-n-signal 9: kin (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
r
5 6
2 4
; 1
» ..........-
0.5 1 1.5 2 2.5 3 3.5 4 Afl^T; 5
Fig. 6. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: two-stage resonant filter. Non-dimensional frequency detuning p0 = -0.5. Initial phase shift ofPSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 7. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: two-stage resonant filter. Non-dimensional frequency detuning p = 0.5. Initial phase shift ofPSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
0.5 1 1.5 2 2.5 3 3.5 4 ДЦ,т5 5
Fig. 8. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: two-stage resonant filter. Non-dimensional frequency detuning p0 = -1. Initial phase shift ofPSK-n-signal 9: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
1 _
yi
\ " \
\ . t
t и v \ * и —-Л-----..... t „ -...... -A—
0,5 1 1.5 2 2.5 3 3.5 4 A£V, 5
Fig. 9. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: two-stage resonant filter. Non-dimensional frequency detuning p0 =1. Initial phase shift ofPSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
increase of irregularity of AFR of passband, but it persists even in the absence of irregularity.
2. For small values of the frequency detuning (|p0| < 0.5) "transparency windows" of dependence of error probability in general are preserved, however, changing the conditions of their manifestation and decreasing their sizes (the decrease in the time interval at which they appear) are observed. At large values of frequency detuning or degree of circuit coupling (P = 2.41) the absence of "transparency windows" is observed for considered PSK-n-signals.
3. Changes in error probability with positive values of frequency detuning for PSK-n-signals considered in this paper, when the value of initial phase shift of signal constellation differs from 9jc = %ln is more pronounced as compared to the
case when it takes negative values when value of frequency detuning is large (|p0| > 0.5). For small values of frequency detuning, the opposite is true.
Results of numerical simulation for first fourth symbols are presented for tuned single stage bandpass filter on coupled circuits (fig. 20, 21). From figures, it follows that dependence of error probability for transmission of the first four symbols does not affect on the choice of resolving time. For resonant filters we can conclude the same according to the results [15].
CONCLUSION
1. When LSS is a bandpass filter, the dependence of the error probability of symbol reception caused by ISI contains
Fig. 10. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: tuned single stage bandpass filter on coupled circuits with degree ofcircuit coupling P=l. nitial phase shift ofPSK-n-signal 9jc: k In (solid line); 0 (dot-dashed line). Number of discrete states n: 4(i);8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 11. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output. LSS: tuned single stage bandpass filter on coupled circuits with degree ofcircuit coupling p = 2.41. Initial phase shift ofPSK-n-signal 9 \K/n (solid line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 12. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree ofcircuit coupling p = 1. Non-dimensional frequency detuning p0 = -0.5. Initial phase shift of PSK-n-signal 9 : n/n (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (5); 128 (6)
------ - _. _
\ \ \ 4 \ \ J \ 6 .... \
\l Y 5
\ \ \ '-л -.л^ \_К /
12 3 4 5 6 ДП„т, 8
Fig. 13. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 1. Non-dimensional frequency detuning p0 = 0.5. Initial phase shift of PSK-n-signal : k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 14. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 1. Non-dimensional frequency detuning p0 = 1. Initial phase shift of PSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
0.6
0.4
0.2
0
1
Fig. 15. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 1. Non-dimensional frequency detuning p0 = 1. Initial phase shift of PSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
100
www.h-es.ru
Fig. 16. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage
bandpass filter on coupled circuits with degree of circuit coupling p = 2.41. Non-dimensional frequency detuning p0 = -0.5. Initial phase shift of PSK-n-signal : k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 17. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 2.41. Non-dimensional frequency detuning p0 = 0.5. Initial phase shift of PSK-n-signal 9: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 18. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 2.41. Non-dimensional frequency detuning p0 = -1. Initial phase shift of PSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (1); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
20
Fig. 19. Error probability per channel symbol depending on channel symbol time duration in the presence of ISI at the LSS output. LSS: single stage bandpass filter on coupled circuits with degree of circuit coupling p = 2.41. Non-dimensional frequency detuning p0 = 1. Initial phase shift of PSK-ra-signal : k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number of discrete states n: 4 (J); 8 (2); 16 (3); 32 (4); 64 (J); 128 (6).
Fig. 20. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output for first four symbols. LSS: tuned single stage bandpass filter on coupled circuits with degree ofcircuit coupling P=l. Initial phase shift ofPSK-n-signal 9jc: k In (solid line); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n = 128. Number ofsymbol: 1 (solid line); 2 (dashed line); 3 (solid line with diamond marker); 4 (solid line with triangular marker); starting from 5 (solid line with circular marker).
additional "transparency windows" at which it is zero, which allows to increase the transmission rate of PSK-n-signals. In the absence of frequency detuning the degree of manifestation of this property increases with an increase of irregularity of AFR of passband, but it persists even in the absence of irregularity. For small values of the frequency detuning "transparency windows" of dependence of error probability in general are preserved, however, changing the conditions of their manifestation and decreasing their sizes are observed.
2. The presence of frequency detuning doesn't lead to change of configuration of resolving time relations between different PSK-n-signals relatively to case of the absence of frequency detuning.
References
1. Proakis J. G. Digital Communications. 4th ed. New York: McGraw-Hill, 2000. 928 p.
2. Zolotarevl. D.,Miller Ya. E. Perekhodnye protsessy v kolebatel'nykh sistemakh i tsepyakh [Transient processes in oscillatory systems and circuits], Moscow: Radiotehnika, 2010. 301 p. (In Russian)
3. FeherK. Wireless digital communications. Modulation and spread spectrum applications. New Jersey: Prentice Hall PTR, 1995. 544 p.
4. Lerner I. M. Method for capacity estimation of real com-municational channels with PSK-n-signals in presence of ISI and its application. T-Comm. 2017. Vol. 11. No. 8. Pp. 52-58.
Fig. 21. Error probability per channel symbol depending on channel symbol time duration in the presence oflSI at the LSS output for first four symbols. LSS: tuned single stage bandpass filter on coupled circuits with degree ofcircuit coupling p = 2.41. Initial phase shift ofPSK-n-signal 9jc: k In (solidline); 0.5k In (dashed line); 0 (dot-dashed line). Number ofdiscrete states n =128. Number ofsymbol: 1 (solidline); 2 (dashedline); 3 (solidline with diamondmarker); 4 (solidline with triangular marker);
starting from 5 (solidline with circular marker).
5. Qureshi S., Newhall E. Adaptive receiver for data transmission over time-dispersive channels. IEEE Trans. 1973. No. 3. Pp. 448-457.
6. Falconer D. D., Magee F. R. Adaptive channel memory truncation for maximum likelihood sequence estimation. Bell System TechnicalJournal. 1973. No. 9. Pp. 1541-1562.
7. Clark A.P., Harvey J.D., Driscoll J.P. Near maximum likelihood detection process for distored digital signals. The Radio andElectronicEngineer. 1978. No. 6. Pp. 301-309.
8. Clark A.P., Najdi N. Y., Fairfield F. J. Data transmission at 19.2 kbits/s over telephone circuits. The Radio andElectron-icEngineer. 1983. No. 4. Pp. 157-166.
9. Vachula G.M., HillF.S.On Optimal Detection of Band-limited PAM Signals with Excess Bandwith. IEEE Trans. 1981. No. 6. Pp. 886-890.
10. Weselowski K. An efficient DFE ML suboptimal receiver for data transmission over dispersive channels using two-dimensional signal constellation. IEEE Trans. 1987. No. 3. Pp. 336-339.
11. ZyablovV. V, Korobkov , PortnoyS.L. Vysokosko-rostnaya peredacha soobshcheniy v real'nykh kanalakh [High-
speed transmission of messages in real channels], Moscow: Radio i svyaz', 1991. 288 p. (in Russian)
12. Il'm G. I., Pol'skii Iu. E. Dinamicheskiy diapazon i toch-nost' radiotekhnicheskikh i optoelektronnykh izmeritel'nykh sistem [Dynamic range and accuracy of radio engineering and optoelectronic measuring systems] Itogi nauki i tekhniki. Ser. Radiotekhnika [The results of science and technology. Radio en-gineeringseries], 1989. Vol. 39. Pp. 67-114. (InRussian)
13. Il'm G. I.Informatsionnye poteri pri prokhozhdenii signalov cherez sistemu s ogranichennoy polosoy propuskani-ya [Information loss during the passage of signals through the system with a limited bandwidth] Vestnik KGTU im. A.N.Tu-poleva [Bulletin of Kazan State Technical University n.a. A.N. Tupolev], 1997. No. 1. Pp. 83-85. (InRussian)
14. Evtyanov S. I. Perekhodnye protsessy v priemno-us-ilitel'nykh skhemakh [Transient processes in the receiver-am-plifiercircuits]. Moscow: Svjaz'izdat. 221 p. (InRussian)
15. Lerner I. M., Il'in G. I. To a question about limition of capacity caused by selective systems with PSK-n-signal impact on it. Nonlinear World. 2017. Vol. 15. No. 1. Pp. 8-12. (In Russian)
К ВОПРОСУ УВЕЛИЧЕНИЯ СКОРОСТИ ПЕРЕДАЧИ В ФАЗОВЫХ РАДИОТЕХНИЧЕСКИХ СИСТЕМАХ ПЕРЕДАЧИ ИНФОРМАЦИИ, РАБОТАЮЩИХ ПРИ СИЛЬНЫХ МЕЖСИМВОЛЬНЫХ ИСКАЖЕНИЯХ В ЛИНЕЙНОМ РАДИОТРАКТЕ
ЛЕРНЕР Илья Михайлович, КЛЮЧЕВЫЕ СЛОВА: повышение скорости передачи;
г. Казань, Россия, aviap@mail.ru межсимвольные искажений; ФМн-п-сигналы; избира-
тельные системы; вероятность ошибки.
КАМАЛЛЕТДИНОВ Наиль Надырович,
г. Казань, Россия, kamal585@mail.ru
АННОТАЦИЯ
Неотъемлемой чертой современного общества является ежегодное увеличение объема информации, который передается посредствам радиотехнических систем передачи информации. При этом закон возрастания объема информации, подлежащей передаче является экспоненциальным. Это приводит к серьезной технической проблеме при создании новых радиотехнических систем передачи информации, поскольку с каждым годом также возрастают требования по эффективному использованию частотного ресурса.
На данный момент одним из наиболее эффективных подходов решения данной проблемы считается переход к передаче, приему и обработке сигналов при наличии межсимвольных искажений в линейных избирательных системах, входящих в линейный радиотракт радиотехнических систем передачи информации. Несмотря на привлекательность такого подхода, его техническая реализация сопряжена с рядом трудностей, которые приводят к увеличению сложности самого приемного устройства при увеличении числа интерферирующих символов. В итоге это приводит к вопросу о практической реализуемости такого устройства.
В данной работе рассматривается вопрос увеличения скорости передачи информации в радиотехнических системах передачи информации, в которых уменьшение межсимвольных искажений в линейных избирательных системах линейного радиотракта не производится, а корректность восстановления передаваемой информационной последовательности достигается за счет соответствующего выбора разрешающего времени. Показано, что при больших значениях объема алфавита канальных символов разрешающее время может быть существенно снижено в том случае, если в качестве линейных избирательных систем используются полосовые фильтры. Это достигается наличие дополнительных «окон прозрачности», при которых вероятность ошибки равна нулю.
СВЕДЕНИЯ ОБ АВТОРАХ:
Лернер И. М., к.ф.-м.н., доцент Казанского национального исследовательского технического университета имени А.Н.Туполева - КАИ;
Камаллетдинов Н.Н., к.т.н., доцент Казанского национального исследовательского технического университета имени А.Н.Туполева - КАИ.
ДЛЯ ЦИТИРОВАНИЯ: Лернер И.М., Камаллетдинов Н.Н. К вопросу увеличения скорости передачи в фазовых радиотехнических системах передачи информации, работающих при сильных межсимвольных искажениях в линейном радиотракте // Наукоемкие технологии в космических исследованиях Земли. 2017. Т. 9. № 5. С. 92-104.
