I have a question that is beyond my nil RF knowledge.
In a situation where a facility is currently running legacy FHSS technologies, what are the considerations to take into account if the plan is to implement DSSS kit alongside it in the same facility?
Due to business reasons the FHSS equipment cant be upgraded or de-commssioned soon, so for a while the two need to co-exist. Is such a setup possible?
desmo, See this tutorial, it may answer your question best.
Jim Geier's article referenced by Reggie is right on the money. FHSS and DSSS, HR/DSSS, and ERP systems do not share the 2.4 GHz ISM band well.
It could have been otherwise. IEEE 802.11 126.96.36.199 defines an optional "Channel Agility" for HR/DSSS and ERP PHYs that, to the best of my knowledge, has never made it into chip sets. Even though protocol analysts regularly see the "Channel Agility" bit set to zero in the Capability Information field of many management frames, it gets no attention and warrants no questions on CWNP Program certification exams.
In brief the idea is that a HR/DSSS Basic Service Set (BSS) would synchronize itself with an overlapping FHSS BSS and hop between three or more of its eleven (USA) channels to avoid being stepped on by the FHSS stations! Sweet.
Again, this feature is probably impossible to find in actual equipment.
I hope this helps. Thanks. /criss
P.S. Here are some clips from the IEEE 802.11-2005-D4 (draft) standards document:
188.8.131.52 Capability Information field: Bit 7 of the Capabilities Information field shall be used to indicate Channel Agility capability by the High Rate direct sequence spread spectrum (HR/DSSS) PHY or ERP. STAs shall set the Channel Agility bit to 1 when Channel Agility is in use, and shall set it to 0 otherwise.
18.1 HR/DSSS PHY Overview: An optional capability for Channel Agility is also provided. This option allows an implementation to overcome some inherent difficulty with static channel assignments (a tone jammer), without burdening all implementations with the added cost of this capability. This option can also be used to implement IEEE 802.11-compliant systems that are interoperable with both FH and DS modulations. See Annex F for more details.
18.1.2: For the purposes of MAC and MAC management, when Channel Agility is both present and enabled (see 18.3.2 and Annex C), the High Rate PHY shall be interpreted to be both a HighRate and a FH PHY.
184.108.40.206 Channel Agility (optional): This Channel Agility option allows an implementation to overcome some inherent difficulty with static channel assignments (a tone jammer), without burdening all implementations with the added cost of this capability. When the Channel Agility option is enabled, the PHY shall meet the requirements on channel switching and settling time, as described in 220.127.116.11, and the hop sequences described below. This option can also be used to implement IEEE 802.11-compliant systems that are interoperable between both FH and DS modulations. Annex F contains a description of the expected behavior when such networks are employed.
18.104.22.168.1 Hop sequences: ...
22.214.171.124 Channel switching/settling time: When the Channel Agility option is enabled, the time to change from one operating channel frequency to another, as specified in 126.96.36.199, is 224 ?s. A conformant PMD meets this switching time specification when the operating channel center frequency has settled to within ???¡À60 kHz of the nominal channel center. Stations shall not transmit until after the channel change settling time.
Annex F (informative)
High Rate PHY/FH interoperability: The Channel Agility option described in 188.8.131.52 provides for IEEE 802.11 FH PHY interoperability with the High Rate PHY. The FH patterns, as defined within this annex, enable synchronization with an FH-PHYcompliant BSS in North America and most of Europe. In addition, CCA requirements on a High Rate station using this mode provide for CCA detection of 1 MHz wide FH signals within the wideband DS channel selected. FH PHY stations operating in mixed mode FH/DS environments are advised to use similar cross PHY CCA mechanisms. The FH (Channel Agility) and cross CCA mechanisms provide the basic mechanisms to enable coexistence and interoperability.
The MAC elements include both DS and FH elements in beacons and probe responses when the Channel Agility option is turned on. Added capability fields indicate the ability to support the Channel Agility option and to indicate whether the option is turned on. These fields allow synchronization to the hopping sequence and timing, identification of what modes are being used within a BSS when joining on either High Rate or FHSS sides, and rejection of an association request in some cases.
Interoperability within an infrastructure BSS can be achieved, as an example, using a virtual dual AP. A virtual dual AP is defined, for purposes of discussion, as two logically separate APs that exist within a single physical AP with a single radio (one transmit and one receive path). Both FHSS and High Rate logical APs send out their own beacons, DTIMs, and other nondirected packets. The two sides interact in the sharing of the medium and the AP?¡é?€??s processor and radio. Addressing and association issues may be handled in one of several ways and are left as an implementation choice.
Minimal interoperability with a nonhopping High Rate or legacy DSSS is provided by the use of a channel at least 1/7 or more of the time. While throughput would be significantly reduced by having a channel only 1/7 of the time, connection and minimal throughput can be provided.
F.1 Additional CCA recommendations: When the FH option is utilized, the HR/DSSS PHY should provide the CCA capability to detect 1 MHz wide FH PHY signals operating within the wideband DS channel at levels 10 dB higher than that specified in 184.108.40.206 for wideband HR/DSSS signals. This is in addition to the primary CCA requirements in 220.127.116.11. A timeout mechanism to avoid excessive deferral to constant CW or other non-IEEE 802.11 type signals is allowed.
FH PHY stations operating in mixed environments should provide similar CCA mechanisms to detect wideband DSSS signals at levels specified in 18.104.22.168, but measured within a 1 MHz bandwidth. Signal levels measured in a full DSSS channel will be generally 10 dB or higher.
It might be possible to define a custom hopping pattern for the legacy FHSS system that avoids the channel(s) used by the DSSS, HR/DSSS, and ERP systems. This is just a guess.
I hope this helps. Thanks. /criss
I seem to remember reading somewhere that a DSSS STA would take about a 10% hit alongside a 2.4 Ghz FHSS STA. But I decided to find out for myself. So I conducted a highly technical experiment here at home . . .
Be careful if trying this experiment by yourself. Otherwise, leave it to the professionals(!) - he he
I have an 802.11g network here at home. (Linksys AP upstairs in the kitchen, and a PCI card in one of my desktop boxes)! I also have a 2.4 Ghz FHSS telephone. Now here's the technical part -- while downloading a ton of data from the Internet and seeing how fast it was downloading ( www.speakeasy.net/speedtest ), I turned that little phone on and off multiple times while downloading multiple times. I also had the telephone antenna right up next to my DSSS client antenna.
Turns out there were times it looked like I took a 200 Kbs hit in download speed, but most of the time, I didn't see any change in download speed what so ever. Oh yeah, I was talking to myself on the phone while doing the test and the phone never seemed to miss a beat either.
So maybe in your interim, both systems CAN operate side by side. If you think about it, a single DSSS channel is 22 Mhz wide, but the FHSS is only using 1 Mhz at any point in time. Statistics have the FHSS STAs operating outside of your DSSS channel to begin with. But when the FHSS STA is using a 1 Mhz channel inside your current DSSS channel, it still isn't doing too much damage. The DSSS data is still getting through for the most part. And the FHSS signal is so strong, that both ends of its link will likely still hear each other just fine too.
So go for it. Besides, if it doesn't work, you can always consider 802.11a technology instead.
By the way, that phone I was using is a brand new Uniden. I got a pair of them with $199/year VOIP service from www.sunrocket.com So far, so good.
First up, thanks for the excellent information.
Compughter - that sums it up nicely, exactly what I was looking for.
Criss - incredible detail, way above what I was expecting but nonetheless very interesting to read. No doubt I will refer back to that content sometime soon. I will investigate the idea of modifying the hopping sequences.
wk4u - great real world example! makes me a whole lot more comfortable about the idea.
If using 802.11a is an option, this would probably be the easiest solution to implement. Most vendors now offer 802.11abg cards. Good luck.
It's highly possible that the Uniden FHSS phone you were using for your test was "Wi-Fi Friendly" which means it employs adaptive frequency hopping (especially if it's only a year old) and thus, would avoid the .11b/g channels in use, using the other FHSS channels for transmit and receive. So, while this is a great feature to have, your test might not give you a real-world example of what will likely happen in an environment where lots of older (or just plain old) FHSS barcode reader/scanners exist. I'm not saying both systems can't coexist but both will incur throughput hits if they are trampling all over each others' channels.
Hmmm. Thanks for the great input Joel. I never thought about that. Kinda blew my highly technical experiment to shreds! BUT I understand what you're talking about.
Nice to learn that newer FHSS equipment will be courteous and listen before transmitting.
Joel (et all).
Turns out my recent experiment wasn't exactly correct. The original question in this thread asked about co locating FHSS STA's and DSSS STA's. NOT OFDM STA's! When looking at the throughput drop, I forgot to switch my AP to the B only mode. My tests were comparing OFDM and FHSS. I thought there was something fishy, but I couldn't quite figure it out until recently. Probably while dreaming in middle of the night.
So I just tried it all over again, only now I'm really comparing FHSS and DSSS. Sure enough, on average, the download speed of a remote station using DSSS takes (complex mathematical equations used here) anywhere from about a 15 to 20% hit in throughput. The FHSS STA's still work just fine. And I'm not sure if this new FHSS phone really does listen for other activity before transmitting on any particular frequency within 2.4 Ghz?
This does go to show that OFDM really is more resilient to interference than DSSS. So like Paul Harvey says, ?¡é?€??now you know the rest of the story?¡é?€??.