Introduction to WirelessWireless TechnologiesInterferenceNetwork SecurityGlossaryBibliography
|
Spread SpectrumBasics - Spread Spectrum - Error Correction - Inter-LANOne of the most common modern methods for reducing the effects of interference is an approach broadly termed spread spectrum (SS). Before spread spectrum, radio broadcasts were tightly focused transmissions centered at a particular frequency. During World War II, the US Navy found itself up against German ships and submarines capable of jamming the Allies' radio-guided torpedoes (and many other military transmissions). Surprisingly, Hollywood actress Hedy Lamarr, along with a pianist, invented a solution: skip between channels fast enough that the enemy would not be able to keep up. They used a player-piano-like roll of punched tape, read in concurrently by both the transmitter and receiver, to describe a particular frequency hopping sequence. The military never paid much attention to their invention, but in the late 1950's engineers rediscovered the approach and it became a critical element of secure communications at the time.
Frequency-hopping Spread Spectrum (FHSS)Lamarr's approach was a primitive form of what is today known as frequency-hopping spread spectrum. For FHSS, a given bandwidth (range of frequencies) is divided into many different distinct channels. Bluetooth, for example, uses the 2.4 - 2.843 Ghz bandwidth, with 79 defined channels. In order to send a signal, the transmitter and receiver must first synchronize on a particular channel and seed value. The transmitter then starts sending its message, only broadcasting for a few hundred milliseconds on any given channel. The transmitter jumps between channels based upon a pseudorandom algorithm; since the receiver knows the starting channel and has the same seed, it can use the same algorithm to follow the jumps and receive the signal. To any device that does not know the jumping sequence, the message is nearly indecipherable. There can be many different FHSS devices broadcasting at the same time in the same area, because when each device is using its own sequence the chance of two of them being on the same channel at the exact same instant is relatively low. FHSS does not significantly affect any tight band transmissions, because there is only an occasional spike of interference on any given channel. And 'bad' channels with a lot of interference only cause a small percentage error in a FHSS transmission, since a single bad channel means only one bit out of 79 will be disrupted.
FHSS has its disadvantages, though. It has a relatively low transfer limit, since only so much information can be sent over any given frequency (remember the Nyquist relationship). There is also no built in redundancy or error checking, which means that once there is a certain critical limit of bad channels, FHSS becomes nearly unusable - too many bits come out corrupted. Direct Sequence Spread Spectrum (DSSS)Because of these disadvantages, recent IEEE 801.11 incarnations (including a, b, and g) replace FHSS with direct sequence, a different form of spread spectrum. DHSS spreads the signal over a larger bandwidth than needed, sacrificing bandwidth efficiency for transmission speed and redundancy. For example, 802.11b DSSS uses 22 Mhz channels, as opposed to the roughly 1 Mhz channels used by FHSS. DSSS accomplishes its frequency spreading by transforming each bit into a distinct, longer sequence. For example, a single '1' bit might become a particular 11-bit sequence (00010011100). At first glance, it might seem that this would greatly reduce transmission speed - after all, you are sending 11 times as much data - but, in reality, the use of many different frequencies makes DSSS much faster than FHSS.
Speed is only one of the advantages of DSSS. The post-processing of a DSSS signal greatly reduces the effects of impulse interference. DSSS signals, by definition, contain a great deal of redundancy, allowing for complex error checking on the received signal. In addition, there is less power required for the transmission on any given single frequency.
Unfortunately, due to DSSS's liberal use of available bandwidth, there is a sharp limit on the number of DSSS users in any given area. For example, with 802.11b's 2.4 - 2.843 Ghz bandwidth, only 3 DSSS users (each using different 22 Mhz ranges) can share the airwaves before encountering interference. |
