O N S PATIAL M ODULATION AND S PACE S HIFT K EYING

P. Maheswaran
Assistant Professor
Dept. of ECE
NIT Trichy.

[email protected]

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 1 / 48

Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 2 / 48

Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 2 / 48

Conventional Vs. Single-RF MIMO

Figure 1: Conventional multi-RF MIMO.1

Figure 2: Single-RF MIMO.

1
A. Mohammadi and F. M. Ghannouchi, “Single RF front-end MIMO transceivers,” IEEE Commun. Mag., vol. 49, no. 12,
pp. 104–109, Dec. 2011, ISSN: 0163-6804.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 3 / 48
Why Single-RF?

The main drawbacks of conventional MIMO systems are2 :

Inter-Channel Interference (ICI): Caused by coupling of multiple symbols in space - in-
creases signal processing complexity.
Inter-Antenna Synchronization (IAS): Detection of symbols at receiver needs all symbols be
transmitted at the same time.
Multiple RF chains: Increasing RF chains causes circuit mismatch, coupling. Expensive to
implement.
Energy Consumption: The energy efficiency decreases linearly with number of RF chains.
2
M. D. Renzo, H. Haas, and P. M. Grant, “Spatial modulation for multiple-antenna wireless systems: A survey,” IEEE Commun.
Mag., vol. 49, no. 12, pp. 182–191, 2011.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 4 / 48
Spatial Modulation: A Single-RF Transmitter
Antenna Switching Unit

MPSK/ Fading
MQAM Channel

Figure 3: Multiplexing in Spatial Modulation

Single RF chain needed as only one symbol is transmitted.

In Spatial Modulation (SM), bit stream from source is divided into log2 (M) +
log2 (Nt ) blocks.
log2 (Nt ) - bits are used to select antennas.
log2 (M) - bits are used to select constellation symbols.
How SM Works?3

3
Renzo, Haas, and Grant, see n. 2.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 6 / 48
How SM Works?4

4
Renzo, Haas, and Grant, see n. 2.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 7 / 48
Space Shift Keying
Space Shift Keying (SSK) is a special form of SM.
In SSK, the input bit stream is mapped to the index of transmit antennas.
The spectral efficiency of SSK is log2 (Nt ), where Nt is no. of transmit
antennas.

10 11 01 00
SSK Modulation Fading
Channel

Figure 4: SSK Transmission

Space Shift Keying
Space Shift Keying (SSK) is a special form of SM.
In SSK, the input bit stream is mapped to the index of transmit antennas.
The spectral efficiency of SSK is log2 (Nt ), where Nt is no. of transmit
antennas.

10 11 01 Fading
SSK Modulation Channel

Figure 5: SSK Transmission

Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 9 / 48

Background

- Diversity gain
BER
- Coding gain

Signal power used as the tower

For a given signal power, system with d2 provides less BER.

For a given BER, system with d2 takes less power to achieve the
BER.
Having larger diversity gain is desirable in a communication sys-
tem.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 10 / 48
Background
Alamouti’s OSTBC5
Introduced a space time code for two antenna transmitters that gives second order
transmit diversity and single symbol decodability.

Spatial Modulation6
Introduced spatial modulation (SM) as a technique to avoid ICI and IAS while maintain-
ing high spectral efficiency. Also introduced an iterative decoding technique for SM.

STBC-SM7
Combines spatial modulation and Alamouti’s space time block code. Information sym-
bols are also expanded into spatial domain (the on-off status of transmit antennas).

5
S. M. Alamouti, “A simple transmit diversity technique for wireless communications,” IEEE J. Sel. Areas Commun., vol. 16,
no. 8, pp. 1451–1458, Oct. 1998.
6
R. Mesleh, H. Haas, C. W. Ahn, and S. Yun, “Spatial modulation-A new low complexity spectral efficiency enhancing
technique,” in First International Conference on Communications and Networking in China, 2006., IEEE, Oct. 2006, pp. 1–5.
7
E. Basar, Ü. Aygölü, E. Panayirci, and H. V. Poor, “Space-time block coded spatial modulation,” IEEE Trans. Commun.,
vol. 59, no. 3, pp. 823–832, Mar. 2011.
Background
CSTSK8
Extends linear dispersion code to spatial modulation to strike a flexible multiplexing and
diversity trade-off.

CIOD-SM9
Uses Complex Interleaved Orthogonal Design meant for two transmit antenna system
in spatial modulation to achieve second order transmit diversity.

TOSD-SSK10
Transmission of signals with good time correlation properties is exploited by transmit
antennas to achieve transmit diversity.

8
S. Sugiura, S. Chen, and L. Hanzo, “Coherent and differential space-time shift keying: A dispersion matrix approach,” IEEE
Trans. Commun., vol. 58, no. 11, pp. 3219–3230, Nov. 2010.
9
R. Rajashekar and K. V. S. Hari, “Modulation diversity for spatial modulation using complex interleaved orthogonal design,”
CoRR, vol. abs/1204.4073, 2012. [Online]. Available: http://arxiv.org/abs/1204.4073.
10
M. D. Renzo and H. Haas, “Space shift keying (SSK-) MIMO over correlated rician fading channels: Performance analysis
and a new method for transmit-diversity,” IEEE Trans. Commun., vol. 59, no. 1, pp. 116–129, Jan. 2011.
Background

Figure 6: Alamouti’s Transmit Diversity Scheme

Transmitter: Receiver:
r r  
Es Es H s
y1 = [h1 s1 + h2 s2 + n1 ], H
r=H y= H H 1 + HH n,
2 2 s2
r r
Es
 
[−h1 s∗2 + h2 s∗1 + n2 ], Es X 2 s1
y2 =
2 r= |hi | I2 + HH n.
2 s2
  r      i=1,2
y1 Es h1 h2 s1 n1
⇒ ∗ = + ∗ .
y2 2 h∗2 −h∗1 s2 n2
 
s
⇒y=H 1 +n
s2
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 13 / 48
Background

Figure 7: Performance comparison of Alamouti and VBLAST schemes

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 14 / 48
Objectives and Motivation

The objective is to find novel methods to improve the performance

of single RF chain multi antenna transmitters like SSK and SM.

Specifically, ways to attain transmit diversity are sought for SSK

and SM.

Existing techniques that improve the performance of SM or SSK

adapt methods invented for multi stream MIMO.

Motivated by this fact, we find methods to achieve transmit diversity

in SSK.

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 15 / 48

Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 15 / 48

Model of a Wireless MIMO System

Multi Antenna Multi Antenna

Fading Channel
Transmitter Receiver

The signal received at the multi antenna receiver is modeled as

y = Hx + n, (1)
where,
I H - Nr × Nt complex matrix, signifies Rayleigh fading.
I x - Nt × 1 vector of transmitted symbols.
I n - Additive white Gaussian noise.
Entries hj,i are complex random variables distributed according to
CN (0, 1).
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 16 / 48
A Property of Complex Normal Random Variables
Exploited for Transmit Diversity

It is proved11 that hi1 − hi2 and hi1 + hi2 are statistically independent
random vectors when i1 6= i2 .

The instantaneous BER of a MIMO system is given based on Q-

function.

The Q-function takes hi1 − hi2 as argument when

 T
0 0 ··· 1 ··· 0 0
x= ↑ (2)
ath position

The Q-function takes hi1 + hi2 as argument when

 T
x = 1 1 ··· 1 ··· 1 1 (3)
11
P. Maheswaran and M. D. Selvaraj, “Performance analysis of feedback-based dynamic SSK-BPSK system,” IEEE Wireless
Commun. Lett., vol. 5, no. 1, pp. 96–99, Feb. 2016.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 17 / 48
Antenna Activation Pattern
for Different Transmit Vectors

Antenna
Selection

Baseband RF Chain

Output power =

1 Input

Figure 8: Antenna activation for x in Eq.(2)

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 18 / 48

Antenna Activation Pattern
for Different Transmit Vectors

Output power =

Power
Baseband RF Chain
Divider

Output power at
each port =
Input

Figure 9: Antenna activation for x in Eq.(3)

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 19 / 48

Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 19 / 48

Feedback Based Transmit Diversity System

Time
Dynamic
Successive MRF-TSSM
SSK-BPSK
SSK-MPSK

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 20 / 48

Dynamic SSK-BPSK

RF Chain

Output power =

In this system model, we consider a single RF chain two antenna

transmitter.

Transmit power of the RF chain is constrained to Γ.

We further assume that the transmitter has a dynamic control unit

that enables it to switch the modulation between SSK and BPSK.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 21 / 48
Dynamic SSK-BPSK: SSK Mode
When SSK is used at the transmitter to send an information symbol
a, the transmitter antenna i = a is activated where a ∈ {1, 2}.
√
The received signal in this case is ys = Γha + n (using (2) in (1)).
The ML detection criterion of 2 × Nr SSK is12

√
 !H 
 Γ 
â = arg max < ys − ha ha (4)
a∈{1,2}  2 

The conditional BER of 2 × Nr SSK is given as

r !
Γ
P1e (||h2 − h1 ||) = Q ||h2 − h1 ||2 . (5)
2
12
J. Jeganathan, A. Ghrayeb, L. Szczecinski, and A. Ceron, “Space shift keying modulation for MIMO channels,” IEEE Trans.
Wireless Commun., vol. 8, no. 7, pp. 3692–3703, Jul. 2009, Eq. (1).
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 22 / 48
Dynamic SSK-BPSK: BPSK Mode
BPSK symbols are notated by sa ∈ {−1, 1} for a ∈ {1, 2}.
q
The received signal is given as yb = Γ2 sa (h1 + h2 ) + n (using (3)
in (1)).

The ML detection criterion for 2 × Nr BPSK can be derived as

â=2
< (h1 + h2 )H yb ≷ 0.

(6)
â=1

The conditional BER of 2 × Nr BPSK can be derived based on (6)

as
q 
2 2
Pe (||h2 + h1 ||) = Q Γ||h2 + h1 || . (7)

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 23 / 48

Dynamic SSK-BPSK: Feedback

Since the receiver has perfect channel state information, using (5)
and (7), we can try to minimize the instantaneous BER as

SSK
P1e (||h2 − h1 ||) ≶ P2e (||h2 + h1 ||) (8)
BPSK

Q(·) is a monotonically decreasing function. Using this, we simplify

(8) as

BPSK
||h2 − h1 ||2 ≶ 2||h2 + h1 ||2 . (9)
SSK

Decision taken by receiver using (9) is fed back to the transmitter

using an ideal feedback channel.

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 24 / 48

Dynamic SSK-BPSK: System Model13

Binary
data SSK/BPSK SSK/BPSK
Modulator ML Detector

Dynamic
Control Modulation Selection

Ideal feedback channel

Figure 10: Dynamic SSK-BPSK System Model. Goto PBPSK

13
Maheswaran and Selvaraj, “Performance Analysis of Feedback-Based Dynamic SSK-BPSK System,” see n. 11.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 25 / 48
Dynamic SSK-BPSK: Performance14
0
10

−1
10

−2
10
BER

−3
10

−4
10 2x2 BPSK
2x2 SSK
2x2 DSB Simulation
−5 2x2 DSB Theory
10
2x2 DSB Asymptotic
2x2 Alamouti Scheme
−6
10
0 5 10 15
SNR(dB)

Figure 11: BER Performance of DSB versus SNR (Nr = 2).

14
Maheswaran and Selvaraj, “Performance Analysis of Feedback-Based Dynamic SSK-BPSK System,” see n. 11.
DSB in Transmitter Correlated Non-identical Fading15
1/2
(a) (b)
1. H = GRt Σ.
 
1 r
2. Rt = .
r 1
 
σ1 0
3. Σ = .
0 σ2
(c) (d)
E[h∗j,1 hj,2 ]
4. r , σ1 σ2 = cos(α).

5. ρ is given as

E[(hj,1 − hj,2 )∗ (hj,1 + hj,2 )]

ρ=
σd σs
2
σ − σ2 2
Figure 12: Vectorial Representation of Random Variables = 1 = cos(θ).
σd σs
15
P. Maheswaran and M. D. Selvaraj, “Dynamic SSK-BPSK system under transmitter correlated non-identical rayleigh fading,”
IEEE Syst. J., May 2018. DOI: 10.1109/JSYST.2018.2828220.
DSB in Transmitter Correlated Non-identical Fading16
1

0.9
2
σ1 = 1, Simulation
0.8 σ2 = 1, Theory
1
2
σ1 = 1.5, Simulation
0.7
σ21 = 1.5, Theory
0.6 σ21 = 1.75, Simulation
PBPSK

σ21 = 1.75, Theory

0.5
2
σ1 = 1.85, Simulation
0.4 σ21 = 1.85, Theory
2
σ1 = 1.9, Simulation
0.3
σ21 = 1.9, Theory

0.2 σ2 = 1.99, Simulation

1
σ21 = 1.99, Theory
0.1
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
r

Figure 13: Probability of choosing BPSK PBPSK for Nr = 2. Goto DSB

16
Maheswaran and Selvaraj, “Dynamic SSK-BPSK System under Transmitter Correlated Non-identical Rayleigh Fading,” see
n. 15.
DSB in Transmitter Correlated Non-identical Fading17
10 -1
r = -0.7, Simulation
r = -0.7, Asymptotic
10 -2 r = -0.3, Simulation
r = -0.3, Asymptotic
r = 0, Simulation
10 -3 r = 0, Asymptotic
r = 0.3, Simulation
r = 0.7, Simulation
BER

10 -4

10 -5
σ12 = σ22 = 1

10 -6

10 -7
0 5 10 15 20
SNR (dB)

Figure 14: Impact of transmitter correlation r on the performance of DSB.

17
Maheswaran and Selvaraj, “Dynamic SSK-BPSK System under Transmitter Correlated Non-identical Rayleigh Fading,” see
n. 15.
DSB in Transmitter Correlated Non-identical Fading18
10 -1
2 2
2x2 BPSK, σ1 = 1.9999, ρ = 0.9998
2x2 DSB, σ 2 = 1.9999, ρ2 = 0.9998
1
10 -2 2 2
2x2 DSB, σ1 = 1.9, ρ = 0.81

2x2 DSB, σ2 = 1.7, ρ2 = 0.49

1
2 2
10 -3 2x2 DSB, σ1 = 1.5, ρ = 0.25

r=0 2x2 DSB, σ12 = 1.3, ρ2 = 0.09

2x2 DSB, σ12 = 1, ρ2 = 0
BER

-4
10

10 -5

10 -6

10 -7
0 5 10 15 20
SNR (dB)

Figure 15: Impact of ρ2 on the diversity of DSB.

18
Maheswaran and Selvaraj, “Dynamic SSK-BPSK System under Transmitter Correlated Non-identical Rayleigh Fading,” see
n. 15.
Outline

1 Introduction

2 Background of Transmit Diversity Systems

3 Novel Single RF Chain Transmit Diversity System: Concept

4 Feedback Based Transmit Diversity System

5 Open Loop Transmit Diversity Systems

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 30 / 48

Open Loop Transmit Diversity Systems

Time
Dynamic
Successive MRF-TSSM
SSK-BPSK
SSK-MPSK

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 31 / 48

Open Loop Transmit Diversity Systems

Time
Dynamic
Successive MRF-TSSM
SSK-BPSK
SSK-MPSK

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 31 / 48

Time Successive SSK-M-ary Modulation (TSSM)

Figure 16: Time Successive SSK-M-ary Modulation

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 32 / 48

How TSSM Achieves Transmit Diversity?

The Pairwise Error Probability of TSSM conditioned on H is derived as

s 
2
 
Γ d
P(a → â|H) = P(Da > Dâ |H) = Q  v1 + aâ v2  , (10)
2 Nt

Es
Where Q(·) is the standard Gaussian Q-function, Γ , daâ , |sa − sâ |. No ,
P 2
Nt
We define random variables (RVs) v1 , ||ha − hâ ||2 and v2 , i=1 hi .

It has been proven that v1 and v2 are independent chi-square RVs19 .

Due to the sum of two independent random variables, TSSM achieves

second order transmit diversity.

19
P. Maheswaran and M. D. Selvaraj, “Time successive SSK-MPSK: A system model to achieve transmit diversity,” IEEE
Commun. Lett., vol. 19, no. 9, pp. 1496–1499, Sep. 2015, ISSN: 1089-7798.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 33 / 48
BER of TSSM20

4x2, QPSK, TSSM Simulation

−1 4x2, QPSK, TSSM Theory
10 4x2, QPSK, TSSM Asymptotic
2x2 TOSD−SSK
CSTSK(2,2,2,2), BPSK
−2 2 Time Slot 4x2 4−PSK
10
2 Time Slot 4x2 SSK
1x2 BPSK
−3
2x2 SSK
BER

10

−4
10

−5
10

−6
10
0 5 10 15 20 25
SNR (dB)

Figure 17: BER performance of various 1 bpcu schemes versus SNR (dB) with Nr = 2.

20
Maheswaran and Selvaraj, “Time Successive SSK-MPSK: A System Model to Achieve Transmit Diversity,” see n. 19.
BER of TSSM21

−1
10

−2
10

−3
BER

10
16x2, 16−PSK, TSSM Simulation
16x2, 16−PSK, TSSM Theory
−4 16x2, 16−PSK, TSSM Asymptotic
10
4x2 TOSD−SSK
CSTSK(2,2,2,4), QPSK
2 Time Slot 16x2 16−PSK
−5
10 2 Time Slot 16x2 SSK
1x2 QPSK
4x2 SSK
−6
10
0 5 10 15 20 25
SNR (dB)

Figure 18: BER performance of various 2 bpcu schemes versus SNR (dB) with Nr = 2.

21
Maheswaran and Selvaraj, “Time Successive SSK-MPSK: A System Model to Achieve Transmit Diversity,” see n. 19.
TSSB in Non-Identical Fading

Time Slot 2 Time Slot 1

BPSK SSK

Detection Metric of TSSB

Figure 19: System Model of TSSB

sa ∈ {−1, 1} for a ∈ {1, 2} - BPSK Constellation.

hi - Nr × 1 Complex Gaussian Random Vector for i ∈ {1, 2}.
hj,i - jth row of hi with CN (0, σi2 ) distribution.
hj1 ,i and hj2 ,i are independent for j1 6= j2 .
Vectorial Interpretation of Random Variables

Figure 20: Random variables as vectors

For identically distributed channel, σ12 = σ22 .

When the channels h1 and h2 are identically distributed, h1 +h2 and
h1 − h2 are independent as they are orthogonal.
Non-identical fading, orthogonality of h1 ± h2
Work-5

Figure 21: Random variables as vectors

For non-identically distributed channel, σ12 6= σ22 .

The orthogonality of h1 +h2 and h1 −h2 is lost in non-identical fading
channel.
(σ 2 −σ 2 )
Their correlation is given as22 ρ = (σ12 +σ22 ) .
1 2
22
P. Maheswaran and M. D. Selvaraj, “On the transmit diversity of time successive SSK-BPSK system,” in Communication
(NCC), 2016 Twenty Second National Conference on, IEEE, 2016, pp. 488–492.
BER of TSSB in Non-identical Fading

Zπ/2 2 2 2 −Nr
1 Γ2 2 (σ1 + σ2 ) 3Γ (σ12 + σ22 )
Pe = (1 − ρ ) + +1 dφ. (11)
π 8 sin4 φ 4 sin2 φ
0

When ρ2 → 1, (11) can be asymptotically approximated as

Γ−Nr 22Nr −1 (2Nr − 1)!!

Pe ≤ . (12)
3Nr (σ12 + σ22 )Nr (2Nr )!!

When ρ2 = 0, (11) can be asymptotically approximated as

Γ−2Nr 23Nr −1 (4Nr − 1)!!

Pe ≤ . (13)
(σ12 + σ22 )2Nr (4Nr )!!
BER of TSSB in Non-identical Fading

−1
10

−2
10

ρ2 = 0 Simulation
−3
10 ρ2 = 0 Theory
ρ2 = 0 Asymptotic
BER

ρ2 = 0.0625 Simulation

−4 ρ2 = 0.0625 Theory
10 2
ρ = 0.5625 Simulation
ρ2 = 0.5625 Theory
2
ρ = 0.7225 Simulation
−5 2
10 ρ = 0.7225 Theory
2
ρ = 0.9801 Simulation
ρ2 = 0.9801 Theory
2
ρ = 0.9801 Asymptotic
−6
10
0 5 SNR (dB) 10 15

Figure 22: BER of TSSB versus SNR (dB) with Nr = 2 for various values of ρ2 .
Multi RF Chain Generalization of TSSM23
Application of TSSM to conventional MIMO transmitter.

Figure 23: Multi RF Chain TSSM

23
P. Maheswaran and M. D. Selvaraj, “Multi RF chain time successive space shift keying-M-ary modulation: A transmit diversity
scheme,” IEEE Trans. Veh. Technol., vol. 66, no. 8, pp. 7086–7097, Aug. 2017. DOI: 10.1109/TVT.2017.2672970.
Multi RF Chain Generalization of TSSM

The signal at the Nr antenna receiver during the first time slot is
NRF
r !
Es X
y1 = hkak + n1 , (14)
P
k=1

Where,
I hki denotes the Nt (k − 1) + ith column of Nr × Nto channel matrix H.
I n1 is the Nr × 1 AWGN introduced by the receiver.
I Entries of hki and n1 are distributed according to CN (0, 1) and
CN (0, No ) respectively.
During the second time slot, the received signal is denoted as
NRF
r Nt
! !
Es X X
k
y2 = hi sak + n2 . (15)
Nt P
k=1 i=1

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 42 / 48

Diversity and Coding Gain of MRF-TSSM

Using (14) and (15), we formulate the codeword matrix of MRF-

TSSM.

We define Nt × 1 vectors eak and o as

 T
1 0 0 ··· 1 ··· 0 0
eak , √ ↑ (16)
P ak th position

1  T
o, √ 1 1 ··· 1 ··· 1 1 . (17)
Nt P

The scaling coefficients in the vectors eak and o follow from (14) and
(15) to take into account the power constraints of the MRF-TSSM
transmitter.

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 43 / 48

Diversity and Coding Gain of MRF-TSSM

Using (16) and (17), the transmitted codeword matrix of MRF-TSSM

transmitter can be given as
 
ea1 sa1 o
 ea sa2 o 
 2
Xa =  .. ..  . (18)

 . . 
eaNRF saNRF o

The error matrix ∆ between the codeword matrices Xa and Xâ

can be obtained from (18) as
 
ea1 − eâ1 (sa1 − sâ1 )o
 ea − eâ (sa2 − sâ2 )o 
2 2
∆ = Xa − Xâ =  .. .. . (19)
 
 . . 
eaNRF − eâNRF (saNRF − sâNRF )o

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 44 / 48

Diversity and Coding Gain of MRF-TSSM

Defining dak âk , (sak − sâk ) and [·]H as the conjugate transpose of a
matrix, we find ∆H ∆ from (19) as
 NRF NRF

T (e − e ) To
P P
(e
k=1 ak − e âk ) ak âk d (e
ak âk ak − eâk ) 
 k=1  (20)
 NP RF NRF 
∗ T 2 T
P
dak âk o (eak − eâk ) |dak âk | o o
k=1 k=1

We see that eTak eâk = (1/P)δak âk and oT o = 1/P, where δak âk is the
Kronecker delta function.
1
Further, eTak o = eTâk o = √
P Nt
which yield the anti-diagonal elements
of (20) as

(eTak o − eTâk o) = (oT eak − oT eâk ) = 0. (21)

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 45 / 48

Diversity and Coding Gain of MRF-TSSM

Using these results, (20) becomes

 NRF 
1 P
P (2 − 2δak âk ) 0
H
 k=1

∆ ∆= 
NRF
.
 (22)
1
|dak âk |2
P
0 P
k=1

We see from (22) that the rank of the matrix ∆H ∆ is 2 if a 6= â.

From this we infer that MRF-TSSM achieves transmit diversity or-

der of 2.

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 46 / 48

Comparison with Transmit Diversity Schemes24

−1
10

−2
10

−3
10
BER

−4
10

2x3, 16−QAM, Alamouti

CSTSK(4,3,2,16), 16−QAM
−5
10 16x3, TOSD−SSK
16x3, 16−QAM, NRF=2 MRF−TSSM
4x3, 8−QAM, Low DoSM CIOD−SM
−6
10
0 2 4 6 8 10 12 14 16 18 20
SNR (dB)

Figure 24: Performance comparison with 4 bpcu transmit diversity schemes.

24
Maheswaran and Selvaraj, see n. 23.
P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 47 / 48
 Queries...

Thank You!

P. Maheswaran, NIT Trichy Spatial Modulation March 20, 2022 48 / 48