802.11b Physical Layer |
Originally the 802.11 standard defined three
physical (PHY) layer standards. These are infra-red and two spread spectrum
systems - frequency hopping spread spectrum (FHSS) and direct sequence spread
spectrum (DSSS). No products have really been developed for infra-red possibly
because of its limited applications. For one thing it is affected by solid
objects and interference from other light sources such as sunlight and
fluorescent lighting.
Spread spectrum
Spread spectrum is a modulation method whereby a pseudo-random or pseudo noise (PN) sequence is used to modulate a carrier signal for transmission. Spread spectrum modulation is used to spread the bandwidth required to transmit a signal [1]. The spreading may be achieved by phase modulation - phase shift keying (PSK) or by varying the carrier frequency - frequency shift keying (FSK). Spread spectrum signals [3] are used for:
FHSS
In frequency hopping a
pseudo-random sequence is used in conjunction with binary or M-ary frequency
shift keying (FSK) to select the frequency of the transmitted signal
pseudo-randomly [3]. The transmitter and receiver pair agree on a hopping
sequence which is used to transmit and recover the information at each end
respectively. In 802.11 the ISM band of 2.4 - 2.4835 GHz is split into 75 1MHz
sub-channels and the hopping pattern is a particular sequence of these channels.
Due to FCC restrictions on bandwidth, data rates were limited to 2Mbps. Thus,
802.11b specifies DSSS exclusively.
It should be noted though, that the FCC recently changed some of the rules
governing frequency hopping systems allowing an increase in data rates to 10
Mbps (see Competition).
DSSS
In direct sequence spread spectrum a PN sequence is used to spread the
information rate of the data signal. The PN sequence is a series of pulses
called chips with typically a much larger bandwidth (chip rate) compared to the
information signal. The ratio of the chip rate to the bit duration of the
information signal is known as the spreading gain and is a measure of the
system's robustness to interference. A detailed treatment of spread spectrum is
provided in [3].
The ISM band is divided into 14 overlapping 22MHz channels with center frequency spacings of 5MHz. Typically in a multiple cell topoplogy, overlapping and/or adjacent cells using different channels can operate simultaneously without interference if the center frequencies are at least 25MHz apart [2]. Channels 1, 6 and 11 are selected for such multicell operation. Table 1 shows the applicable channel assignments for different geographical regions. Channel 14 is designated specifically for Japan.
Table 1 High rate PHY frequency channel plan [2]
CHANNEL ID |
Center frequency |
FCC (USA) |
IC (Canada) |
ETSI (Europe) |
Spain |
France |
MKK (Japan) |
1 |
2412 |
X |
X |
X |
- |
- |
- |
2 |
2417 |
X |
X |
X |
- |
- |
- |
3 |
2422 |
X |
X |
X |
- |
- |
- |
4 |
2427 |
X |
X |
X |
- |
- |
- |
5 |
2432 |
X |
X |
X |
- |
- |
- |
6 |
2437 |
X |
X |
X |
- |
- |
- |
7 |
2442 |
X |
X |
X |
- |
- |
- |
8 |
2447 |
X |
X |
X |
- |
- |
- |
9 |
2452 |
X |
X |
X |
- |
- |
- |
10 |
2457 |
X |
X |
X |
X |
X |
- |
11 |
2462 |
X |
X |
X |
X |
X |
- |
12 |
2467 |
- |
- |
X |
- |
X |
- |
13 |
2472 |
- |
- |
X |
- |
X |
- |
14 |
2484 |
- |
- |
- |
- |
- |
X |
The original 802.11 standard specified an 11 bit PN code known as a Barker sequence for 1 and 2 Mbps data rates. Each 11 bit sequence represents a single data bit (1 or 0) and is transmitted at a symbol rate of 1 Msps using DBPSK. For better spectral efficiency DQPSK is used to achieve 2 Mbps.
In order to achieve higher data rates 802.11b specifies the use of a more advanced coding technique called complementary code keying (CCK) which was proposed by Lucent and Harris. It has been shown to possess good Euclidean distance properties in a multipath environment [4] which is advantageous in a wireless medium. Complementary codes were first discovered by M.J.E. Golay for infrared multislit spectrometry. They possess some unique properties that make them ideal for communications.
Given the ith code word xl = [x1,
x2,........,
xN]T
the autocorrelation is
given by
A complementary pair has the property [4] that
Notice that there is just one peak at zero shift. In practice this ideal property may not be obtained but there should be one main peak at zero shift with all other peaks much smaller.
There are also polyphase codes in existence. 802.11b specifies a set of 64 8-bit complex codewords given as:
Table 2. Phase parameter encoding table [2,5]
Dibit |
Phase parameter |
Dibit value |
Phase |
(d1, d0) |
j1 |
00 | 0 |
(d3, d2) |
j2 |
01 |
p |
(d5, d4) |
j3 |
10 |
p/2 |
(d7, d6) |
j4 |
11 |
-p/2 |
8 data bits are used to form a code word. As
shown in Table 2 the first dibit specifies j1 and the value of the bits
determine the phase value and likewise for the other three dibits. Notice in the
code that j1 appears in all the elements and is said to merely rotate the code.
In an implementation for 11 Mbps by Intersil [5], 6 bits are used to choose one
of the set of 64 code words while the other two are used to QPSK modulate a
carrier. A detailed analysis of the CCK implementation is provided in [5]. For
5.5 Mbps 4 bits are used to modulate the carrier.
The complete specifications are given in Table 3 for all modulation types
Table
3 802.11b Data Rate Specifications [1]
Data Rate (Mbps) |
Code Length |
Modulation |
Symbol Rate (Msps) |
Bits/Symbol |
1 |
11 (Barker) |
DBPSK |
1 |
1 |
2 |
11
(Barker) |
DQPSK |
1 |
2 |
5.5 |
8 (CCK) |
DQPSK |
1.375 |
4 |
11 |
8 (CCK) |
DQPSK |
1.375 |
8 |
Theoretically users communicate at 11 Mbps
speeds but this is not practical just as in a wired ethernet. The best reported
speed are about 7.5 Mbps. As distance between devices or to the access point
increases and/or in noisy fading environment, the data rate also decreases.
802.11b provides for a dynamic rate shifting whereby the data rate may be
reduced to 5.5, 2 or 1 Mbps based on the channel or distance conditions.
This rate shifting is purely a physical layer mechanism transparent to the
higher layers.