 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention pertains to satellite communication systems and more
particularly is concerned with spread spectrum or code division multiple
access satellite communication systems.
There are at present very small aperture terminal (VSAT) networks which
employ code division multiple access (CDMA) techniques. These networks,
however, do not permit distinct networks to share the same inbound space
segment.
There is a need for private satellite networks containing between 20 and
about 300 terminals in communication with a common hub ground station.
Code division multiple access (CDMA) spread spectrum techniques were first
used in military systems because of the cost of their implementation.
Further, the possibility has been raised of systems with spread spectrum
techniques in combination with time slotted packet contention techniques
such as that known as ALOHA.
SUMMARY OF THE INVENTION
Briefly, there is provided a satellite communication system having a hub
earth station and a plurality of terminal earth stations. Each inbound
satellite communication link from a terminal earth station to the hub
earth station carries digital data packets encoded in CDMA code for code
division multiple access at a first bit rate during synchronized time
slots on a contention channel. An outbound communication link from the hub
earth station to the terminal earth stations sends time division
multiplexed digital data at a second bit rate higher than the first bit
rate. Another aspect of the invention forms a network with a number of
systems sharing the same hub, each system using a different CDMA code.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a diagram of a satellite system embodying the invention; and
FIG. 2 illustrates the communication links between a hub and terminal of
the system of FIG. 1.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
In keeping with the invention, code division multiple access (CDMA) spread
spectrum techniques are used in satellite communication systems and
networks. The invention allows the sharing of an inbound space segment in
a non-interfering manner by a number of terminal earth stations belonging
to several distinct networks, while at the same time avoiding the
necessity for shared hub facilities. Each network includes a corresponding
hub ground station.
Referring to FIG. 1, a space satellite 10 provides shared relay service for
a plurality of distinct networks A, B. Each network includes a plurality
of terminal ground stations 11 and a corresponding hub ground station 12.
The terminal ground stations transmit packets to their respective hubs
using the same inbound frequency bands. Each terminal ground station
employs direct sequence spread spectrum transmission using a code in the
form of pseudorandom binary sequence, e.g. a Gold Code, to spread the
spectrum of a data-modulated binary phase shift keyed (BPSK) signal. This
results in a transmitted signal of the form:
s(t)=Ac(t)d(t)coswt.
where c(t) is a binary-valued waveform at rate 1/T.sub.c representing the
spread sequence and d(t) is the binary-valued data waveform at rate 1/T.
In all cases, T is much greater than T.sub.c. Also, the data waveform,
d(t), may include r=1/2, K=7 convolutional encoding. The ratio T/T.sub.c
is known as the spreading factor or processing gain of the spread spectrum
system.
A network, which is defined as a plurality of terminals coupled to a
corresponding hub, may include several systems. All the terminals of a
given system employ the same pseudorandom sequence code to spread their
transmitted spectra. Different systems of a network use different codes,
but share the same hub.
Each network has its own hub transmitting a time division multiplexed (TDM)
outbound link at a bit rate higher than that of an inbound link. To meet
FCC flux density limitations and to minimize the required antenna size at
the terminals, some spectral spreading may be used in the outbound links
from the hub. Despreading at the terminals may be accomplished using a
corresponding baseband despreader.
In the inbound direction, from terminal to hub, each network utilizes a
slotted multiaccess contention protocol, e.g. ALOHA. Slot synchronization
is provided within each network; no attempt is made to synchronize slots
among the different networks sharing the inlink. Collisions of packets
within each individual system cause the destruction of all the colliding
packets as the same spreading code is used. Each individual system can
operate at a throughput of about 0.20 packets/slot, corresponding to an
offered load of about 0.26 packets/slot. Because each individual system
uses a different spreading code, collisions between the packets of
different systems are not destructive but merely cause an increase in the
effective ambient noise level. Thus as the number of systems sharing the
inlink increases, performance degrades gracefully. For a spreading ratio
of T/T.sub.c =127, between 5 and 10 distinct spreading codes or signature
sequences can simultaneously occupy the bandwidth with less than 1 dB of
performance degradation, assuming that all the signatures are present all
the time and that r=1/2, k=7 convolutional encoding is employed. Because
of the low per-system traffic load, between 15 and 20 small to medium size
networks are able to share the inbound bandwidth. This number may be
increased by increasing the spreading ratio.
In a CDMA environment, one of the more difficult problems is that of
acquiring code synchronization at the receiver so that despreading and
demodulation can be successfully accomplished. This usually requires a
combination of a matched filter and code tracking loop. In the present
case because of the slotted environment, code synchronization is inherent
within each network as the code starts on the slot boundary. Therefore,
the receivers at the hubs may be simple active correlators followed by
phase shift demodulators. However, if there is large tolerance in the slot
boundaries, matched filters may be needed.
The use of coherent phase shift keying on the inlinks maximizes the
performance of the convolutional code, and therefore, as noted earlier,
maximizes the multiple access capability of the overall inlink. This has
the added benefit of allowing the minimization of the transmit power from
the terminals.
As indicated above, each individual network employs a slotted contention
protocol on the inlink to its hub. Each system of a network employs a
unique nominal signature or spreading sequence, so that all other systems
that share the frequency band appear as noise. This is further aided by
the fact that each network is self-synchronous, but is asynchronous to all
others. The overlap of two or more signals from terminals in the same
system, and therefore having the same spreading code, appears as a
collision. All packets involved in the collision will be lost and will
have to be retransmitted. This hard collision can occur only among the
packets belonging to a particular system. Collisions between packets from
different systems, and therefore having different spreading codes, will
lead only to a small increase in the ambient noise level and therefore, to
a minor increase in the rate at which packets are corrupted by bit errors.
Provided the penalty due to the CDMA is small (less than 1 dB), then the
throughput of any of the N systems sharing the band may be expressed
approximately as:
S=Ge.sup.-G packets/slot
where G is the offered load in packets per slot. This is the classical
slotted ALOHA expression for throughput. On the other hand, if N systems,
each with throughput S packets per slot and slot duration T, are occupying
the band, then the overall channel utilization or throughput may be
expressed as NS packets per slot-time. Because of the CDMA transmission
format, this can be greater than one. If the per-system throughput is set
at S=0.20 and NS=5, 25 systems may share the inlink bandwidth. Note that a
system is a group of terminals that share a given spreading code as well
as a shared hub. A given network may require more than one nominal
spreading code, i.e., it may have too many terminals for one inbound
channel. This may be accommodated by adding an additional receive chain at
the hub for that network. However, each additional nominal spreading code
is regarded as an additional system using the frequency band.
EXAMPLE
Suppose that terminals transmit at a data rate of 4,800 bits/sec and that
r=1/2, K=7 convolutional encoding is employed. If each terminal
transmission is spread by a factor of 127, each transmission is at a
channel rate of 1,219,200 chips and a channel bandwidth of about 3 MHz
will be required.
Further suppose that each terminal network consists of 100 terminals and
that each such network operates at a throughput of S=0.20 packets/slot,
corresponding to an offered load of B=0.26 packets/slot. If on average,
there are 5 packets simultaneously in the channel from different networks,
then the same channel can support approximately N=5/0.2=25 simultaneous
networks with an overall performance penalty of about 1/2 dB in received E
/N. This allows a channel to support an overall population of around 2,500
terminals.
It will be appreciated that the invention provides a code division multiple
access approach for allowing a large population of terminal terminals,
organized into a moderately large group of small to medium-size networks,
to occupy the same inbound space segment in an essentially non-interfering
manner. Each network has its own hub and operates using a slotted
contention protocol on the inlink and a TDM stream in a separate channel
on its outlink.
Provision is made to have distinct private networks share the same inbound
space segment or radio-frequency channel. This is done by assigning each
system of a network a unique nominal signature or spreading code.
The invention is particularly suitable for small to medium-size terminal
networks that each have a private hub rather than the shared hub that is
usually required when more than one network shares inbound space segment.
* * * * *
|
|
 |
|
 |
Description |
|
