Tech Stuff - Frequency Hopping
vs
Direct Sequence
First, lets start out by saying that each technology has its strengths and weaknesses. It depends on your needs. This is our biased (no really!!) overview of 802.11 technolgies.
ED Note: This piece was originally written before the arrival of 802.11g and ODFM and has been updated since this technology has a significant effect on the efficiency of DSSS links.
A quick overview of each technology.
Frequency Hopping Spread Spectrum (FHSS): derived from military radio technology where it was designed to be inherently secure and reliable under adverse battle conditions. Divides the available 83.5 MHz spectrum (in most countries) into 79 (or 75) discrete 1 MHz channels (the 4.5 MHz left over provides a 'guard bands' at either end of the spectrum), the Radio then hops around these 1 MHz channels in a pseudo-random sequence, using a minimum of 75 frequencies every 30 seconds and using any single frequency for a max. of 400 milliseconds.
Direct Sequence Spread Spectrum (DSSS): designed originally by two vendors to increase the available speed on the wireless network. Divides the available 83.5 MHz spectrum (in most countries) into 3 wide-band 22 MHz channels. Uses an 11 bit spreading code to reduce the possible interference on signals in each wide-band channel. Because of sensitivity to interference a very wide (22MHz) frequency range DHSS must have method of progressively falling back to lower throughput rates and indeed these do go as low as 1Mb.
Wireless performance is a function of how much bandwidth is available in each channel. DHSS which uses 22 MHz channels will always be faster that FHSS which uses 1 MHz channels both in the UNII/ISM bands. Though by using techniques such as channel bonding FHSS can be used to reach 3 - 5 Mb range. As a rule of thumb modulation techniques will give 0.5 to 2+ times the MHz in bit/s throughput. So a 22 MHz using CCK gives ~11Mb (0.5) throughput and using OFDM (11.g) with is giving 55Mb (2+).
A note about the effect of Orthogonal frequency-division multiplexing (OFDM) OFDM splits the available 22Mhz DSSS channel into a series of smaller channels - somewhat analogous to DSL on wireline networks. The theory being that any one of the smaller channels may suffer from multipath or fading problems but not all of them at the same time. If enough get through even in seriously congested areas then transmission will continue. Pure shut-out (as is possible with 11.b) is almost impossible with OFDM. In this sense OFDM exhibits many of the properties of FHSS. Simply put, many of our objections to DSSS at the gross level have been significantly ameliorated by ODFM.
Warning: Depends what you want from your technology. We wanted the most cost-effective, reliable wireless communication we could find - not necessarily the fastest. We'll accept a customer can suffer performance problems from time to time to due adjacent wireless and other radiating installations BUT NEVER a shut-out. Based on this point-of-view read the rest of this piece at your discretion. The use of (D)PCF which seems to be the default mode in most systems significantly ameliorates the shut-out effect for DSSS at the cost of throughput.
Reliable wireless performance is a combination of:
- Transmission Power
- Antenna design and diversity
- Range
- Noise
We control the amount of radio power we can cost-effectively supply (in our case a high 200mW). We can and do supply very cost-effective antennas. We can advise customers on the range and antennas to be used. We cannot control the 'noise'. So for 'outdoor' operations we believe 'noise' is the key element in the equation.
As wireless becomes more popular (and that is inevitable) we must start from the following worst case assumptions:
- It's very noisy out there
- It's going to get worse
Note: Pretending the problem disappears at 5.8 GHz is at best lunacy. The relative lack of 'noise' in this spectrum is a transient and temporary phenomena.
So let's play out the nightmare scenario with the two technologies.
Lets suppose our neighbor turns up a powerful (full 4 watt EIRP) DSSS radio within 100 meters. If you had been using a DS radio quite happily on the same channel you would now be shut out - 100%. Zero throughput. The FH radio would lose about 28% throughput (not good but not the end-of-the-world). So you reconfigure your DS radio to use another channel and you walk into another DS radio blasting away. Dead again. Zero throughput. Our FH radio is still operating (yes, we've lost 55% throughput) and we haven't done anything yet. So our DS radio is now moved to the 3rd channel and ... you've guessed, we get hit again. Zero throughput. Our only solution - Move location.
Our intrepid FH radio is now down to 25% throughput but still running. And we still haven't done anything to improve the situation.
So our choice of radio technology boils down to this. If you can control absolutely all the interference and noise between your stations - use Direct Sequence technology (in-building, airports etc). If you can't - use Frequency Hopping.
But, what about the loss of performance you will say. DS is typically quoted as 11 Mbps (actually its not) whereas FH is around 1 - 3 Mbps. Guess what, 11 Mbps x nothing = zero Mbps.
In the real world of outdoor ISP access and network solutions Frequency Hopping is practical.
Problems, comments, suggestions, corrections (including broken links) or something to add? Please take the time from a busy life to 'mail us' (at top of screen), the webmaster (below) or info-support at zytrax. You will have a warm inner glow for the rest of the day.
