IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO.

4, APRIL 2003 937

Super-Orthogonal Space–Time Trellis Codes

Hamid Jafarkhani, Senior Member, IEEE, and Nambi Seshadri, Fellow, IEEE

Abstract—We introduce a new class of space-time codes called method to maximize the coding gain for a given rate, constella-
super-orthogonal space-time trellis codes. These codes combine tion, and number of states.
set partitioning and a super set of orthogonal space–time block In [7], Alamouti introduced a simple code to provide full di-
codes in a systematic way to provide full diversity and improved
coding gain over earlier space–time trellis code constructions. versity for two transmit antennas. In [8], the scheme is general-
We also study the optimality of our set partitioning and provide ized to an arbitrary number of antennas and is named space–time
coding gain analysis. Codes operating at different rates, up to block coding. Also, the theory of orthogonal designs has been
the highest theoretically possible rate, for different number of generalized in [8] to show when it is possible to achieve full
states can be designed by using our optimal set partitioning. diversity. Although a space–time block code provides full di-
Super-orthogonal space-time trellis codes can provide a tradeoff
between rate and coding gain. Simulation results show more than versity and a very simple decoding scheme, despite the name,
2-dB improvements over the codes presented in the literature its main goal is not to provide the additional coding gain [8],
while providing a systematic design methodology. [9]. This is in contrast to space–time trellis codes, which pro-
Index Terms—Orthogonal designs, set partitioning, space–time vide full diversity as well as coding gain but at a cost of higher
codes, super-orthogonal codes, transmitter diversity, trellis codes. decoding complexity. To achieve additional coding gain, one
should concatenate an outer code such as a trellis code with an
inner space–time block code. In [10], space–time block coding
I. INTRODUCTION is combined with a trellis code to provide more coding gains.
The same scheme is used in [11] for Rayleigh-fading channels
S PACE–TIME trellis codes have been introduced in [1] to
provide improved error performance for wireless systems
using multiple transmit antennas. The authors have shown that
with large space–time correlations and simulation results are
provided. The shortcoming of the scheme in [10] and [11] is the
such codes can provide full diversity gain as well as additional fact that it does not provide the highest possible rate. The idea
signal-to-noise ratio (SNR) advantage that they call the coding of concatenating a space–time block code with an outer trellis
gain. Code design rules for achieving full diversity are also pro- code is also exploited in [12], [13].
vided. Using these design rules, examples of codes with full di- In what follows, we combine space–time block codes with
versity as well as some coding gain were constructed that are not a trellis code to come up with a new structure that guarantees
necessarily optimal. Since there is no general rule for designing the full diversity with increased rate over [10] and [11]. Also,
codes that provide diversity as well as coding gain, it is unclear we show how to design the trellis code to maximize the coding
how to design new codes for different number of states or dif- gain. The result is a systematic method to design space–time
ferent rates. Also, it is not clear how to improve the performance trellis codes for any given rate and number of states. To the best
of the codes, i.e., how to maximize the coding gain. There have of our knowledge, this is the first systematic way of designing
been many efforts to improve the performance of the original space–time trellis codes. Not only do we show how to design a
space–time trellis codes [2]–[5]. While very interesting codes space–time trellis code for a given rate and number of states, but
have been proposed in the literature, the coding gain improve- also our general set-partitioning results provide the maximum
ments are marginal for one receive antenna. In this work, not coding gain for the proposed structure. During the review of
only do we propose a scheme that improves the performance by this manuscript, we realized that Siwamogsatham and Fitz have
more than 2 dB, but also we answer the questions of systematic independently come up with similar ideas [14]–[16]. One spe-
design for any rate and number of states and the maximization cific example of our general super-orthogonal space–time trellis
of the coding gain. We provide a new structure for space–time codes is also presented in [17] independently. Part of the current
trellis codes that guarantees full diversity and provides opportu- work has been presented as conference papers in [18] and [19].
nity to maximize the coding gain. We also provide a systematic Section II provides the motivation behind this work. In Sec-
tion III, first we provide a parameterized class of space–time
block codes. Then, we study the set partitioning of space–time
block codes using phase-shift keying (PSK) constellation sym-
Manuscript received November 14, 2001; revised September 2, 2002. The
material in this paper was presented in part at the International Communications bols. Using the new parameterized class of space–time block
Conference, New York, April 2002 and at the IEEE International Symposium codes and our set partitioning, we show how to design optimal
on Information Theory, Lausanne, Switzerland, June 2002. super-orthogonal space–time trellis codes. Section IV provides
H. Jafarkhani was with Broadcom Corporation, Irvine, CA 92620 USA. He
is now with the Department of Electrical and Computer Engineering, University an analysis of the coding gain for the super-orthogonal codes. In
of California, Irvine, Irvine, CA 92697 USA (e-mail: [email protected]). Section V, we extend our systematic method of designing super-
N. Seshadri is with Broadcom Corporation, Irvine, CA 92620 USA (e-mail: orthogonal space–time trellis codes to more than two transmit
[email protected]).
Communicated by G. Caire, Associate Editor for Communications. antennas. We present simulation results in Section VI. Finally,
Digital Object Identifier 10.1109/TIT.2003.809607 some concluding remarks are provided in Section VII.
0018-9448/03$17.00 © 2003 IEEE
938 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

II. MOTIVATION using two of the codes in . For example, one can generate the
following four matrices using the code in (1):
An example of a full-rate full-diversity complex space–time
block code is the scheme proposed in [7] which is defined by
the following transmission matrix:
(3)
(1)
There are four other possible distinct orthogonal matrices
which are listed as follows:
The scheme can be used for transmit antennas and any
number of receive antennas. The scheme transmits bits every
two symbol intervals, where the two–dimensional (2-D) con-
stellation size is . For each block, bits arrive at the (4)
encoder and the encoder chooses two modulation symbols To create these additional matrices, one can use the following
and . Then, using , the encoder transmits from code from the set :
antenna one and from antenna two at time one. Also, the en-
coder transmits from antenna one and from antenna two (5)
at time two. This scheme provides diversity gain, but no addi-
tional coding gain.
which represents a phase shift of the signals transmitted from
By concatenating an outer trellis code that has been designed
antenna one by . We denote a set including all orthogonal
for the additive white Gaussian noise (AWGN) channel with the
matrices from (3) and (4) as . It is important to note that
space–time block code, additional performance gain can be ob-
the rank of a matrix (which determines diversity) based on the
tained. To see this, view each of the orthogonal matrices
difference between any two distinct matrices within either (3)
generated by the space–time block code (1) as a four–dimen-
or (4) is , but the rank of a matrix obtained by considering the
sional (4-D) signal point (strictly speaking while there are four
difference between any two elements in (3) and (4) is .
elements in the matrix, it does not create a 4-D space and there
By using more than one code from set , we can create all pos-
is only two degrees of freedom). The outer trellis code’s task is
sible orthogonal matrices from . Therefore, the scheme
to select one of the 4-D signal points to be transmitted based on
provides a sufficient number of constellation matrices to design
the current state and the input bits. In [10], it is shown that for
a trellis code with the highest possible rate. Also, it allows a
the slow fading channel, the trellis code should be based on the
systematic design of space–time trellis codes using the avail-
set partitioning concepts of “Ungerboeck codes” for the AWGN
able knowledge about trellis-coded modulation (TCM) [20] and
channel.
multiple TCM (MTCM) [21] in the literature.
One shortcoming of the scheme proposed in [10] and in [11]
In the space–time trellis codes of [1], constellation sym-
is the fact that there is a rate loss associated with achieving any
bols are assigned to each trellis branch. So, choosing a trellis
coding gain if the constituent 2-D signal constellation size does
branch is equivalent to transmitting symbols from transmit
not increase. This is because these schemes are not using all of
antennas in one time slot. What is transmitted at the next time
the possible 4-D signal constellations.
slot depends on the next selected trellis branch and is not deter-
To elaborate, consider other codes which provide behavior mined automatically. The codes in [1] are designed such that
similar to those of (1) for the same rate and number of transmit a maximum diversity and rate are guaranteed. However, it is
antennas. The set of all such codes which only use , , and not clear if the highest possible coding gain is achieved. In this
their conjugates with positive or negative signs are listed as fol- work, we present new codes that not only provide maximum di-
lows: versity and rate, but also achieve coding gains higher than those
of the codes in [1].
In our new scheme, we assign a space–time block code with
specific constellation symbols to transitions originating from a
state. Therefore, in general, for a space–time block code,
picking a trellis branch emanating from a state is equivalent to
transmitting symbols from transmit antennas in time
intervals. By doing so, it is guaranteed that we get the diversity
(2)
of the corresponding space–time block code, as in [10] and [11],
while in what follows we show how to design the trellis code for
With a small abuse of notation, we call the union of all these the highest possible rate, as was done in [1], to get the maximum
codes as “super-orthogonal code” set . Using just one of the coding gain as well. Note that different space–time block codes
constituent codes from , e.g., the code in (1), one cannot create can be assigned to different trellis branches as long as they all
all possible orthogonal matrices for a given constella- provide the same diversity (otherwise, the diversity is defined
tion. To make this point more evident, let us concentrate on bi- by the lowest diversity). We elaborate on the issues of picking
nary phase-shift keying (BPSK) constellation for now. It can be the right space–time block codes and assigning them to different
shown that one can build all possible orthogonal matrices trellis branches in the following sections.
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 939

III. SUPER-ORTHOGONAL CODES

A. A Parameterized Class of Space–Time Block Codes
In this work, we use the following class of orthogonal designs
as transmission matrices:

(6)

Note that provides the code in (1). So, with a slight

abuse of notation, we have . Let us
concentrate on the case for which the constellation signal al-
phabet is not expanded. Note that the transmitted signals for
are , , , and . We pick such
that for any choice of and from the original constella-
Fig. 1. Set partitioning for BPSK; the numbers at leaves represent the indexes
tion points, the resulting transmitted signals are also from the of the symbols in the space–time block code.
same constellation. For example, if we use -PSK constella-
tion signals (PSK constellations containing points), the con-
stellation signals (and thus and ) can be represented by codes and “the expansion of the signal constellation” that is usu-
, . One can pick , where ally a negative side effect. Our super-orthogonal code expands
. In this case, the resulting transmitted sig- the number of available orthogonal matrices and, as we show in
nals are also members of the -PSK constellation and, therefore, the sequel, this is the main reason why we can design full-rate
do not expand the constellation signals. Since the transmitted trellis codes that provide full diversity. It has no negative side
signals are from a PSK constellation, the peak-to-average power effect either in terms of expanding the transmitted signal con-
ratio of the transmitted signals is equal to one. So, not only do stellation or in terms of increasing the peak-to-average power
we not increase the number of signals in the constellation, but ratio.
also there is no need for an amplifier to provide a higher linear
operation region. More specifically, we use and , B. Set Partitioning for Orthogonal Codes
, , for BPSK and quaternary PSK (QPSK), respec- This section provides a set partitioning for orthogonal codes
tively. By using and for the BPSK and shows how to maximize the coding gain. Let us denote the
constellation, one can generate all orthogonal matrices in difference of the transmission matrices for codewords and
. In fact, is the code in (1) and is by and its Hermetian, complex conjugate and trans-
the code in (5). By using (1) and (5), the set of transmitted sig- pose, by . Following the definitions in [1], the diver-
nals consists of , , , , , . For any symmetric sity of such a code is defined by the minimum rank of the matrix
constellation, the set of transmitted signals is the same as the . For a full-diversity code, the minimum of the deter-
set of constellation signals. This includes quadrature amplitude minant of the matrix over
modulation (QAM) constellations as well as PSK constellations. all possible pairs of distinct codewords and corresponds
We call the combination of these two codes a super-orthogonal to the coding gain. We define the coding gain distance (CGD)
code. In general, a super-orthogonal code consists of the union between codewords and as ,
of a few orthogonal codes, like the ones in (6). A special case where is the determinant of matrix . In general, if in-
is when the super-orthogonal code consists of only one orthog- stead of a full diversity we have a code with diversity less
onal code, e.g., only . Therefore, the set of orthogonal than , the distance can be defined as the harmonic mean of
codes is a subset of the set of super-orthogonal codes. Obvi- the nonzero eigenvalues of . Then, we use CGD in-
ously, the number of orthogonal matrices that a super-orthog- stead of Euclidean distance to define a set partitioning similar
onal code provides is more than (or in the worst case equal to) to Ungerboeck’s set partitioning [20].
the number of orthogonal matrices that an orthogonal code pro- Let us concentrate on the case where we only utilize the code
vides. Therefore, while the super-orthogonal code does not ex- in (1). Also, let us assume that we use a BPSK constellation.
tend the constellation alphabet of the transmitted signals, it does Consider a four-way partitioning of the orthogonal code as
expand the number of available orthogonal matrices. This is of shown in Fig. 1 for BPSK. At the root of the tree, the minimum
great benefit and crucial in the design of full-rate, full-diver- determinant is . At the first level of partitioning, the highest
sity trellis codes. Another advantage of super-orthogonal codes determinant that can be obtained is . This is obtained by
lies in the fact that the code is parameterized. An important ex- a set partitioning in which subsets and use different
ample of a super-orthogonal code that we use throughout this transmitted signal elements for different transmit antennas. At
paper is the union of for an -PSK constellation the next level of partitioning, we have four sets , , ,
where , and . Another example and with only one element per set.
is the union of for a QAM constellation where A four-state trellis code (Fig. 2) can be constructed based on
. the set partitioning of Fig. 1 with a rate of bits per second
Note that one needs to separate “the expansion of the orthog- per hertz (bits/s/Hz) for an -PSK constellation where .
onal matrices” that is a positive result of using super-orthogonal This is similar to the codes presented in [10] and [11] which can
940 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

Note that (10) includes a sum of terms and each of these terms
is nonnegative. Therefore, the following inequality holds:

(11)
Based on the coding distances calculated in (9) and (10), one
can show that the coding gain of such a space–time trellis code is
dominated by parallel transitions. The optimal set partitioning
for BPSK, QPSK, and 8-PSK are demonstrated in Figs. 1, 3,
and 4, respectively. As can be seen from these figures, the min-
imum CGD increases (or remains the same) as we go one level
down in the tree. The branches at each level can be used to de-
Fig. 2. A four-state trellis; Ungerboeck’s set partitioning.
sign a trellis code with a specific rate. Higher coding gain ne-
cessitates the use of redundancy resulting in reduced rate. In the
only transmit a rate which is half of the maximum possible rate following sections, we show how to design space–time trellis
[1]. For an -PSK constellation, each signal can be represented codes without sacrificing the rate.
by , or , where .
Now, we consider two distinct pairs of constellation symbols

and
C. Set Partitioning for Super-Orthogonal Codes
and the corresponding code matrices and to calculate This subsection provides a set partitioning for super-orthog-
and . For the sake of brevity, in the sequel, onal codes and shows how to maximize the coding gain without
we omit from and when there is no ambiguity. sacrificing the rate. Code construction based on a super-orthog-
For parallel transitions in a trellis, we have (7) and (8) as shown onal set is as follows. In our new scheme, we assign a constituent
at the bottom of the page. Using (8), one can show that space–time block code to all transitions from a state. The adja-
cent states are typically assigned to one of the other constituent
(9) space–time block codes from the super-orthogonal code. Simi-
larly, we can assign the same space–time block code to branches
Now, if we have two codewords which differ in pairs of con-
that are merging into a state. It is thus assured that any path that
stellation symbols, it can be shown that still .
diverges from (or merges to) the correct path differs by rank
Also, if for the first codeword, we denote the set of constellation
. In other words, every pair of codewords diverging from (or
symbols by
merging to) a state achieves full diversity because the pair is
from the same orthogonal code (same parameter ). On the other
hand, for codewords with different , it is possible that they do
and for the second codeword, we denote the set of constellation not achieve full diversity. Since these codewords are assigned
symbols by to different states, the resulting trellis code would provide full
diversity despite the fact that a pair of codewords in a super-or-
thogonal code may not achieve full diversity. Note that this is
just a general method to guarantee full diversity. It is possible
we have
to come up with examples that do not follow this rule and still
provide full diversity.
Similar to the case of orthogonal designs, it remains to do the
set partitioning such that the CGD is maximized at each level
(10) of partitioning. We use the formulas that we have developed

(7)

(8)
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 941

Fig. 3. Set partitioning for QPSK; the numbers at leaves represent the indexes of the symbols in the space–time block code.

Fig. 4. Set partitioning for 8-PSK; the numbers at leaves represent the indexes of the symbols in the space–time block code.

in Section III-B to calculate the CGDs. Figs. 1, 3, and 4 show D. Super-Orthogonal Space–Time Trellis Codes
the set partitioning for BPSK, QPSK, and 8-PSK, respectively. In this section, we show how to use our proposed set par-
Using (9) to calculate the CGD between a pair of codewords, it titioning scheme to design full-diversity full-rate space–time
is apparent that increasing the Euclidean distance between the trellis codes. First, we start with a few important examples and
first signals of the codewords will increase the CGD. The CGD then we propose some general rules how to design a super-or-
also increases as we increase the Euclidean distance between thogonal space–time trellis code for a given trellis and required
the second signals of the codewords. Therefore, a rule of thumb rate. Figs. 5–10 demonstrate examples of our new super-orthog-
in set partitioning is to choose the codewords that contain signal onal space–time trellis codes. In these figures, rep-
elements with highest maximum Euclidean distance from each resents the particular member of our parameterized space–time
other as the leaves of the set-partitioning tree. For example, in block code which is used at the specific state. Also, we have
the case of QPSK in Fig. 3, , , are the shown the corresponding sets from our set partitioning next to
QPSK signal constellation elements and repre- each state.
sent , respectively. The maximum CGD in this Fig. 5 shows a four-state example of our new super-orthog-
case is when and in (9). This onal space–time trellis codes. In this example, when we use
is the justification for the choice of leaf codewords in Fig. 3. At BPSK and the corresponding set partitioning in Fig. 1, the rate of
the second level of the tree from the bottom, it is impossible to the code is one. We use when departing from states
have both and equal to in all cases. The next zero and two and use when departing from states
highest value for CGD is when , or one and three. Note that, with this new structure, we have eight
, . Therefore, we group the subtrees in possible orthogonal matrices instead of four which allows
the second level such that the worst case is when , us to design a full-rate code. The minimum CGD of this code
or , . We keep grouping is which can be found in Fig. 1 and Table I. In Section IV,
the subtrees to maximize the minimum CGD at each level of set we prove that parallel transitions are dominant in calculating the
partitioning. Similar strategies are used for other signal constel- minimum CGD for this code.
lations. If we use a QPSK constellation and the corresponding set
In Figs. 1, 3, and 4, set partitioning is used to assign signal partitioning in Fig. 5, the result is a four-state super-orthogonal
elements to branches diverging from (or merging to) a state to space–time trellis code at rate 2 bits/s/Hz. The minimum CGD
maximize coding gain. The optimality of these set partitioning for this 2-bits/s/Hz code is equal to which is greater than ,
is discussed in Section IV. the CGD of the corresponding space–time trellis code from [1].
942 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

Fig. 5. A four-state code; r = 1 bit/s/Hz using BPSK or r = 2 bits/s/Hz using

QPSK.

Fig. 6. A four-state code; r = 3 bits/s/Hz (8-PSK). Fig. 8. An eight-state code; r = 3 bits/s/Hz (8-PSK).

We also provide an eight-state, rate 3-bits/s/Hz (8-PSK) ex-

ample in Fig. 8. The CGD of this code is while the CGD
of a similar code in [1] is . To design such a code, we have
some degrees of freedom in choosing the rotations and the sets.
Therefore, there are different code designs that provide the same
coding gain. Another example of an eight-state, rate 3-bits/s/Hz
code with similar properties is provided in Fig. 9. Note that the
Fig. 7. A two-state code; r = 1 bit/s/Hz using BPSK or r = 2 bits/s/Hz using only limitation in picking different options is to avoid a cata-
QPSK.
strophic code. To avoid a catastrophic code, a change of a few
input bits should not create an infinite number of different sym-
Fig. 6 shows a 3-bits/s/Hz super-orthogonal space–time bols. In other words, the same input bits should not create the
trellis code using 8-PSK and the corresponding set partitioning same codeword when starting from different states. To achieve
of Fig. 4. The minimum CGD for this code is equal to . We this goal, either the for the orthogonal code assigned to dif-
study the details of the coding gain calculations in Section IV. ferent states should be different or the assigned subsets should
There is no four-state space–time trellis code for 8-PSK in [1]. be different. As the number of states grows, we can pick be-
Fig. 7 demonstrates the codes for a two-state trellis providing tween these two options (or their combinations) and this is ba-
a minimum CGD of and at rates 1 bit/s/Hz using BPSK sically the main difference between Figs. 8 and 9. This is a
and 2 bits/s/Hz using QPSK, respectively. There are no equiva- good example to show the possibility of having different as-
lent codes reported in [1]. signments in the code. In fact, there are choices of rotations that
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 943

Fig. 10. A four-state code; r = 2.5 bits/s/Hz (8-PSK).

the super-orthogonal code. Similarly, we can assign the same
orthogonal code to branches that are merging into a state. It
is thus assured that any path that diverges from (or merges to)
the correct path is full rank. Then different sets from the corre-
sponding set partitioning are assigned to different transitions
similar to the way that we assign sets in a regular MTCM. To
avoid a catastrophic code, the subsets are assigned such that
for the same input bits in different states either the rotation
angles of the orthogonal codes are different or the assigned
subsets are different. Since the process of set partitioning is
done to maximize the CGD at each level, the coding gain of
the resulting super-orthogonal space–time trellis code is also
maximum.
Note that unlike the manual design strategy in [1], one can
systematically design a code for an arbitrary trellis and rate
using our set partitioning and code design strategy. While we
have shown the examples of full-rate codes, in general, codes
with lower rates can be designed to provide higher coding gains.
Fig. 9. An eight-state code; r = 3 bits/s/Hz (8-PSK). The design method is exactly the same. We only use sets from
a different level of the set partitioning. Using different levels of
the set partitioning to design super-orthogonal space–time trellis
make the overall CGD less than and must be avoided. If a
codes provides a tradeoff between rate and coding gain. For ex-
higher coding gain is required at 3 bits/s/Hz, one should
ample, a four-state rate 2.5 bits/s/Hz code using 8-PSK with a
use a 16-state super-orthogonal space–time trellis code. Similar
CGD of is shown in Fig. 10. The maximum possible rate using
to conventional trellis codes, there is always a tradeoff between
8-PSK is and an example is shown in Fig. 8.
coding gain and the number of states (complexity).
Table I tabulates the CGD of our new super-orthogonal
Codes with different number of states and at different rates
space–time trellis codes and the corresponding codes from
can be systematically designed using the set partitioning in
[1]. It is evident from Table I that using our super-orthogonal
Figs. 1, 3, and 4. The rules for assigning different sets to
space–time trellis coding method, not only can we system-
different transitions in the trellis are similar to the general rules
atically design codes that did not exist before, but also the
of thumb defined in [20] and [21] to design MTCM schemes.
CGD of the new codes are consistently better than those of the
First, based on the required rate, we select a constellation
original codes presented in [1]. In Section VI, we compare the
and use the corresponding set partitioning. The choice of the
performance of our super-orthogonal space–time trellis codes
constellation also defines the valid rotation angles that would
with the best available space–time trellis codes in the literature.
not create an expansion of the transmitted signal constellation.
We use these valid values for in (6) to define the super-or-
thogonal code to be used in the trellis code. Then, we assign a IV. CGD ANALYSIS
constituent orthogonal code from the set of super-orthogonal In this section, we derive the coding gain of different super-
codes to all transitions from a state. The adjacent states are orthogonal space–time trellis codes that we introduced in Sec-
typically assigned to another constituent orthogonal code from tion III. We show how to find the dominant path for CGD calcu-
944 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

TABLE I
CGD VALUES FOR DIFFERENT CODES

(13)
For , for example, the trellis in Fig. 5, we have

(14)
Since , we have

(15)
Fig. 11. Two typical paths differing in P = 3 transitions. Now, we prove the following result that is used in the calcu-
lation of the coding gains.
lation in the trellis. Our approach is general enough to be easily
extended to other trellises in the literature. We first consider spe- For the four-state code in Fig. 5, the minimum value of
cific examples given in Section III and calculate their coding the CGD when is greater than the minimum value
gains. Then, we generalize the methods that we have used in the of the CGD when .
calculation of these coding gains to show how to calculate the Proof: Without loss of generality, we assume two code-
coding gain of any super-orthogonal space–time trellis code. words diverging from state zero and re-emerging after tran-
sitions to state zero. For parallel transitions , one can
A. Error Events With Path Length of Three calculate the CGD from (9). In this case, for
Let us first consider the trellis in Fig. 2 which corresponds to the 1-bit/s/Hz code using BPSK and for
the code in Fig. 5. Note that when there are parallel transitions the 2-bits/s/Hz code using QPSK. For the 1-bit/s/Hz
between two states, we assign a different transition to each pos- code using BPSK, and we have ,
sible constellation matrix (symbols) which is defined by our set , and . Therefore, inequality
partitioning. Two codewords may only differ in trellis (15) results in
transition. However, due to the structure of the trellis, it is im-
possible to have two codewords which differ in two trellis transi- (16)
tions . Because, for example, if two codewords diverge Also, for
from state zero, they have to go through at least three transi-
tions to re-emerge (Fig. 11). Therefore, the smallest value of
excluding parallel transitions is three. For , we consider a and in (12), we have which means
typical case where the first codeword stays at state zero. For the
(17)
second codeword, the first and third transitions (diverging and
merging to state zero) use and the second transi- Combining inequalities (16) and (17) provides
tion uses as in Fig. 11. Using trigonometry equa- (18)
tions, it can be shown that
which is greater than .
(12) For the 2-bits/s/Hz code using QPSK, and
where we have , , and
. Therefore, inequality (15) results in
(19)
Also, for
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 945

TABLE II
MINIMUM det(A) FROM (24) FOR DIFFERENT CONSTELLATIONS AND ROTATIONS

B. Error Events With Path Length of Two

We consider two codewords diverging from state zero and
re-emerging after transitions to state zero in Fig. 12.
For parallel transitions , one can calculate the CGD
from (9). We consider a typical case where the first codeword
stays at state zero. For the second codeword, the first transition
diverging from state zero uses and the second tran-
sition reemerging to state zero uses as in Fig. 12.
Fig. 12. Two typical paths differing in P = 2 transitions.
Note that , , in Fig. 6, in Figs. 7 and 8,
and in Fig. 9. It can be shown that
and in (12), we have which means
(20) (22)

Combining inequalities (19) and (20) shows where

(21)
which is greater than .
Using the preceding result, one can calculate the coding gain
of the codes in Fig. 5. It is easy to show that the minimum value
of the CGD when is greater than the minimum value (23)
of the CGD when . This proves that the minimum CGD
and . Since only depends on ,
for the code in Fig. 5 is dominated by parallel transitions and is
is not negative, and are independent of
equal to for BPSK. Therefore, the coding gain is for 1
, one can first calculate and then cal-
bit/s/Hz using BPSK. Also, the coding gain of the code in Fig. 5
culate . In other words, one can use the following
for 2 bits/se/Hz using QPSK is . This is due to the fact
formula to find :
that the minimum CGD for this code is dominated by parallel
transitions and is equal to . Therefore, the coding gain is .
Note that we have considered specific codes for clarity of (24)
the presentation. One can show similar results for any trellis for We use (24) to calculate the CGD for trellises with . We
which it is impossible to diverge from a state and merge to the tabulate for different constellations and rotations
same state in transitions. For the trellises that contain in Table II. In Table II, we use the notation of Fig. 12, where
a path with transitions (Fig. 12), we provide similar with a slight abuse of the notation ,
formulas in the sequel. , , and . We use these values to
946 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

Fig. 14. Two paths with length four differing in P = 2 transitions.

other paths with providing different CGDs. Therefore,
for the code in Fig. 13, the minimum CGD is not a good indi-
cator of the performance. In addition to the minimum CGD, one
needs to consider the path weights (multiplicity) of error events.
In this case, study of the distance spectrum is required to find
the best code, i.e., the best assignment of rotations and subsets.
Fig. 13. A four-state code; r = 3 bits/s/Hz (8-PSK). Also note that this is not a linear code. While we provide a few
examples, a complete study to find the best codes for each trellis
and rate is beyond the scope of this paper and is left for future
calculate the CGD of super-orthogonal space–time trellis codes. research.
Again we consider specific examples for the sake of simplicity.
First, we calculate the coding gain for the BPSK code in V. EXTENSION TO MORE THAN TWO ANTENNAS
Fig. 7. We assume two codewords diverging from state zero and
re-emerging after transitions to state zero. For parallel transi- In this section, we extend our general approach for designing
tions , one can calculate the CGD from (9) which is . super-orthogonal space–time trellis codes (SOSTTCs) to more
For , and than two transmit antennas. We follow the same principles to
systematically design SOSTTCs using orthogonal designs, con-
stellation rotations, and set partitioning. We provide a few ex-
(25) amples to show how our general approach works.
Therefore, CGD and the coding gain is dominated by
paths with transitions and is equal to . A. Real Constellations
For the QPSK code in Fig. 7, the CGD for parallel transitions A full-rate real orthogonal design only exists for
is . For , since , we have [8]. An example of a real orthogonal design from
[8] is given as follows:
(26)
Therefore, the coding gain for the QPSK code in Fig. 7 is .
In the case of the 8-PSK codes in Figs. 8 and 9, the parallel
(27)
transitions are dominant and the coding gain is CGD .
This can be shown using (24) and Table II.
The coding gain in Fig. 6 is dominated by paths with
transitions. Using (24) and Table II, one can show that the min- To expand the orthogonal matrices, similar to the case of two
imum CGD is as shown in Table I. We emphasis the im- antennas in (6), we use phase rotations as follows:
portance of picking the right set of rotations and subsets in pro-
viding the maximum coding gain. We also note that the min-
imum CGD is not always enough in comparing two codes with
each other. As an example, let us consider the four-state, rate
3-bits/s/Hz code in Fig. 13. We have picked (28)
for states one and three in Fig. 13 instead of
in Fig. 6. The code in Fig. 13 is an interesting example where
starting from state zero, parallel transitions are dominant and the
CGD is . However, one should consider other error events with Since the constellation is real (for example, BPSK or PAM) and
length two where the parallel transitions are not dominant. For we do not want to expand the constellation signals, we pick
example, starting from state one and staying at state one for two , where . This means that we only potentially
transitions, an error event with can be a path diverging use a sign change for each column. Note that, in general, for
from state one to state two and re-emerging to state one in the transmit antennas, one can use rotations. Equation (27)
second transition. Note that this error event can be part of two is a member of the super-orthogonal sets in (28) for
codewords with length four as illustrated in Fig. 14. The CGD . In other words, with a slight abuse of the notation, we
for such a path is which is lower than and . There are have .
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 947

Fig. 15. Set partitioning for four transmit antennas (BPSK); the numbers at leaves represent the indexes of the symbols in the space–time block code.

Similar to the case of two transmit antennas, we provide a set

partitioning for super-orthogonal codes and show how to maxi-
mize the coding gain without sacrificing the rate. Following the
definitions in [1], for a full-diversity code, the minimum of the
determinant of the matrix
over all possible pairs of distinct codewords and corre-
sponds to the coding gain. Similar to the case of two transmit
Fig. 16. A two-state code; four transmit antennas; r = 1 bit/s/Hz (BPSK).
antennas, the CGD between codewords and is defined as
, where is the determinant
of matrix . Then, we use the CGD instead of Euclidean dis-
tance to define a set partitioning similar to Ungerboeck’s set
partitioning [20]. An example for a BPSK constellation is shown
in Fig. 15. At the root of the tree, the minimum determinant is
. At the first level of partitioning, the highest determinant
that can be obtained is , which is obtained by creating sub-
sets and with transmitted signal elements differing at least
in two positions. At the last level of partitioning, we have eight
sets , , , , , , , and with two
elements per set that differ in all positions. The resulting min-
imum CGD is .
Code construction based on super-orthogonal sets is as
follows. We assign a constituent space–time block code to
all transitions from a state. The adjacent states are typically
assigned to one of the other constituent space–time block code
from the super-orthogonal code. Similarly, we can assign the Fig. 17. A four-state code; four transmit antennas; r = 1 bit/s/Hz (BPSK).
same space–time block code to branches that are merging into
a state. It is thus assured that any path that diverges from (or B. Complex Constellations
merges to) the correct path differs by rank . In other words,
every pair of codewords diverging from (or merging to) a A full-rate complex orthogonal design does not exist for more
state achieves full diversity because the pair is from the same than two transmit antennas [8]. A rate complex orthogonal
orthogonal code (same parameters , , ). On the other design for four transmit antennas was proposed in [8]. It has
hand, for codewords with different , , , it is possible been shown in [22] that this is the maximum possible rate for
four transmit antennas. An example of a rate complex
that they do not achieve full diversity. Since these codewords
orthogonal design is given as follows:
are assigned to different states, the resulting trellis code would
provide full diversity despite the fact that a pair of codewords
in a super-orthogonal code may not achieve full diversity.
Figs. 16–18 show the examples of two-state, four-state, and (29)
eight-state codes for transmitting 1 bit/s/Hz using BPSK,
respectively.
948 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

We compare our results with those of the existing space-time

trellis codes in the literature when a comparable code exists. In
all simulations, similar to the results in [1], a frame consists of
130 transmissions out of each transmit antenna.
Fig. 19 shows the frame error probability results versus SNR
for the codes in Figs. 5 and 7 using BPSK and the corresponding
set partitioning in Fig. 1. Both of these codes are full-rate and
transmit 1 bit/s/Hz. Note that one cannot design such a code
using the scheme developed in [10] and [11].
Fig. 20 shows the simulation results for transmitting 2
bits/s/Hz using a QPSK constellation. The codes in Figs. 5 and
7 are denoted by “SOSTTC 4 states” and “SOSTTC 2 states,”
respectively. The corresponding results for a code with the
same rate and four states from [1] is also provided (“STTC”)
for comparison. We also include the results for four-state codes
form [3] (“YB”), [4] (“BBH”), and [5] (“CYV”). To the best
of our knowledge, these are the best available codes in the
literature. Note that while the codes in [3]–[5] provide some im-
provements over the original space–time trellis codes in [1] for
more than one receive antennas, their performance is very close
to that of [1] for the case of one receive antenna. As can be seen
from the figure, our four-state super-orthogonal space–time
trellis code outperforms the corresponding space–time trellis
codes by more than 2 dB. The performance of our four-state
SOSTTC outperforms that of a 32-state space–time trellis
codes in [1] and is very close to that of a 64-state space–time
trellis code in [1]. We also note that our SOSTTC preserves
the decoding simplicity of orthogonal designs (with added
complexity for the Viterbi algorithm). In fact, when redundant
computations are avoided and the orthogonality of the building
blocks is smartly used, the decoding complexity of a SOSTTC
is less than that of a space–time trllis code using the same
trellis. The details of simple decoding are explained in [14],
[15] and we do not repeat them here.

Fig. 18. An eight-state code; four transmit antennas; r = 1 bit/s/Hz (BPSK). VII. CONCLUSION

We have introduced a new structure for designing space–time

To expand the orthogonal matrices, similar to the case of real trellis codes which is called super-orthogonal space-time trellis
constellations, we use the following rotations: codes (SOSTTCs). While providing full diversity and full rate,
the structure of our new codes allows an increase in the coding
gain. Not only does our SOSTTC outperform the space-time
trellis codes in the literature, but it also provides a systematic
method for designing space–time trellis codes at different rates
(30) and for different trellises. Since we have used orthogonal de-
signs as the building blocks in our SOSTTCs, the complexity
of the decoding remains low while full diversity is guaranteed.
Codes operating at different rates, up to the highest theoretically
Again, we only use rotations that do not expand the constella-
possible rate, for different number of states, can be designed
tion. For example, , , can be for QPSK.
by using our optimal set partitioning. In general, SOSTTCs can
Then, similar to the case of real constellations and two transmit
provide a tradeoff between rate and coding gain while achieving
antennas, after set partitioning we can systematically design
full diversity.
SOSTTCs for any trellis and complex constellation.
Our code design strategy is general enough such that a quasi-
orthogonal space–time block code [23] or any other structure
VI. SIMULATION RESULTS
which guarantees diversity is applicable as well. Although our
In this section, we provide simulation results for our new general approach is applicable to other nonorthogonal coding
SOSTTCs using two transmit antennas and one receive antenna. structures, there are many details that are left for future research.
JAFARKHANI AND SESHADRI: SUPER-ORTHOGONAL SPACE–TIME TRELLIS CODES 949

Fig. 19. Simulation results for r = 1 bit/s/Hz (BPSK).

Fig. 20. Simulation results for r = 2 bits/s/Hz (QPSK).

950 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 49, NO. 4, APRIL 2003

ACKNOWLEDGMENT [10] S. Alamouti, V. Tarokh, and P. Poon, “Trellis-coded modulation and

transmit diversity: Design criteria and performance evaluation,” in Proc.
IEEE Int. Conf. Universal Personal Communications (ICUPC-98), vol.
The authors wish to thank the anonymous reviewers and 2, 1998, pp. 917–920.
Dr. Marvin Simon for their comments that have improved the [11] S. Siwamogsatham and M. P. Fitz, “Robust space-time coding for corre-
presentation of the paper. lated Rayleigh fading channels,” IEEE Trans. Signal Processiong, vol.
50, pp. 2408–2416, Oct. 2002.
[12] M. J. Borran, M. Memarzadeh, and B. Aazhang, “Design of coded mod-
ulation schemes for orthogonal transmit diversity,” in Proc. IEEE Int.
Symp. Information Theory (ISIT), Washington, DC, June 2001, p. 339.
REFERENCES [13] S. Sandhu, R. Heath, and A. Paulraj, “Space-time block codes versus
space-time trellis codes,” in Proc. IEEE Int. Conf. Communications
[1] V. Tarokh, N. Seshadri, and A. R. Calderbank, “Space-time codes for (ICC), vol. 4, June 2001, pp. 1132–1136.
high data rate wireless communication: Performance analysis and code [14] S. Siwamogsatham and M. P. Fitz, “Improved high rate space-time codes
construction,” IEEE Trans. Inform. Theory, vol. 44, pp. 744–765, Mar. via orthogonality and set partitioning,” in Proc. IEEE Wireless Commu-
1998. nications and Networking Conf. (WCNC), Mar. 2002.
[2] J. Grimm, M. P. Fitz, and J. V. Krogmeier, “Further results on space-time [15] , “Improved high rate space-time codes via concatenation of
codes for Rayleigh fading,” in Proc. 36th Annu. Allerton Conf., Sept. expanded orthogonal block code and M-TCM,” in IEEE International
1998, pp. 391–400. Conference on Communications (ICC), vol. 1, Apr. 2002, pp. 636–640.
[3] Q. Yan and R. S. Blum, “Optimum space-time convolutional codes for [16] , “Improved high-rate space-time codes via expanded STBC-
quasistatic slow fading channels,” in Proc. IEEE Wireless Communica- MTCM constructions,” in Proc. IEEE Int. Symp. Information Theory
tions and Networking Conf. (WCNC), vol. 3, Sept. 2000, pp. 1351–1355. (ISIT), Lausanne, Switzerland, June/July 2002, p. 106.
[4] S. Baro, G. Bauch, and A. Hansmann, “Improved codes for space-time [17] M. Ionescu, K. K. Mukkavilli, Z. Yan, and J. Lilleberg, “Improved 8- and
trellis-coded modulation,” IEEE Commun. Letters, vol. 4, pp. 20–22, 16-state space time codes for 4psk with two transmit antennas,” IEEE
Jan. 2000. Commun. Lett., vol. 5, pp. 301–305, July 2001.
[5] Z. Chen, J. Yuan, and B. Vucetic, “Improved space-time trellis coded [18] N. Seshadri and H. Jafarkhani, “Super-orthogonal space-time trellis
modulation scheme on slow fading channels,” Electron. Lett., vol. 37, codes,” in Proc. IEEE Int. Conf. Communications (ICC), vol. 3, Apr.
pp. 440–441, Mar. 2001. 2002, pp. 1439–1443.
[6] A. R. Hammons, Jr. and H. El-Gamal, “On the theory of space-time [19] H. Jafarkhani and N. Seshadri, “Optimal set-partitioning for super-or-
codes for PSK modulation,” IEEE Trans. Inform. Theory, vol. 46, pp. thogonal space-time trellis codes,” in Proc. IEEE Int. Symp. Information
524–532, Mar. 2000. Theory (ISIT), Lausanne, Switzerland, June/July 2002, p. 249.
[7] S. M. Alamouti, “A simple transmitter diversity scheme for wireless [20] G. Ungerboeck, “Channel coding for multilevel/phase signals,” IEEE
communications,” IEEE J. Select. Areas Commun., vol. 16, pp. Trans. Inform. Theory, vol. IT-28, pp. 55–67, Jan. 1982.
1451–1458, Oct. 1998. [21] D. Divsalar and M. K. Simon, “Multiple trellis coded modulation
[8] V. Tarokh, H. Jafarkhani, and A. R. Calderbank, “Space-time block (MTCM),” IEEE Trans. Commun., vol. 36, pp. 410–419, Apr. 1988.
codes from orthogonal designs,” IEEE Trans. Inform. Theory, vol. 45, [22] H. Wang and X.-G. Xia, “Upper bounds of rates of space-time block
pp. 1456–1467, July 1999. codes from complex orthogonal designs,” in Proc. IEEE Int. Symp. In-
[9] , “Space-time block coding for wireless communications: Perfor- formation Theory (ISIT), Lausanne, Switzerland, June 2002, p. 303.
mance results,” IEEE J. Select. Areas Commun., vol. 17, pp. 451–460, [23] H. Jafarkhani, “A quasi-orthogonal space-time block code,” IEEE Trans.
Mar. 1999. Commun., vol. 49, pp. 1–4, Jan. 2001.