Question on using FHSS (single antenna) vs. Using 2 antennas with different frequencies

[PB82]

Hi everyone,

I’m currently studying some diversity techniques in wireless communications, and a question came to mind when studying time/space/frequency diversity.

I would like to compare the pros and cons of 2 options:

• Using a single antenna to do FHSS.
• Using 2 antennas, each in a well separated band (ex: 400MHz and 900MHz).

Here is some context as to why I am interested in this topic:

• I’m an electrical engineering student, and I’m working on a project involving a wireless sensor network. The “router” nodes of this network each have 2 antennas. One operates around 400MHz, one operates around 900 MHz.
• The designers of this device designed it with plenty of redundant components. Nearly all hardware in the node has redundant copies.
• The purpose of making this device was mainly to demonstrate a redundancy management scheme that the team was working on.
• The decision for using antennas operating at different frequencies was arbitrary. There was no theoretical analysis done before making this decision. Therefore, I’m trying to figure out if this is a good decision, or if there are better alternatives.

Here’s what I’ve observed so far in my reading:

• In MIMO cellular systems, I see that the antennas of a MIMO system all operate at one frequency. This is space diversity. Also, this allows for techniques such as beamforming.
• But I wonder, is there any use for having multiple antennas, but at different frequencies?
• I assume this no longer counts as space diversity. This is now frequency diversity. Also, this would eliminate the possibility of beamforming (I think).
• If the difference between the carrier frequencies of the antennas is greater than the coherence bandwidth of the channel, then the frequency selective attenuation of the signals will be independent. To me, this sounds like a good idea, so then why is this not more common?
• Perhaps this is not common because FHSS offers better performance. Or, maybe FHSS doesn’t offer better performance, but it’s cheaper (important for WSNs) and offers ‘good enough’ performance. Or maybe it’s some other practical/theoretical reason like increased security, or some other theoretical/practical consideration.

I’m most interested in insights that are particularly relevant for wireless sensor networks, especially ideas that can be implemented with a long battery life.

However, I’m open to hear about any thoughts you guys have on this. I know this is super open ended, I hope that’s okay. Frankly, I don’t think I know enough to be able to narrow the scope of my question, and I want to hear about all sorts of thoughts that come to your mind. I realize there's probably a million things to consider, so I'd be grateful to hear about them.

Thanks! And let me know if there’s more information you need, I’ll try my best to answer.

 

[prcguy]

With todays technology I would think that one transceiver can be designed that covers two bands like 400MHz and 900MHz including FHSS and dual band antennas are very common. That should save some $$ without sacrificing much performance. The few FHSS voice radios I have used seem to work as well as single frequency radios within the same band so I don’t think you have to give up performance with FHSS.

I think operating in two different frequency bands simultaneously will decrease the possibility of detection and increase data throughput with the increased BW plus it can help mitigate RF interference problems. Both space diversity and frequency diversity can help with multipath and fading problems but that’s usually more important for mobile operations.

Otherwise I’m not a data guy and have more experience with RF voice circuits.

 

[PB82]

With todays technology I would think that one transceiver can be designed that covers two bands like 400MHz and 900MHz including FHSS and dual band antennas are very common. That should save some $$ without sacrificing much performance. The few FHSS voice radios I have used seem to work as well as single frequency radios within the same band so I don’t think you have to give up performance with FHSS.

I think operating in two different frequency bands simultaneously will decrease the possibility of detection and increase data throughput with the increased BW plus it can help mitigate RF interference problems. Both space diversity and frequency diversity can help with multipath and fading problems but that’s usually more important for mobile operations.

Otherwise I’m not a data guy and have more experience with RF voice circuits.

Thank you for the response! I'll start to evaluate some options with dual-band antennas.
I think our device is going to have 2 antennas on it regardless (for the sake of redundancy), but I'm certainly open to changes regarding the types of antennas and frequency bands, so I'll definitly take a look at some dual-band options.
In regards to using 2 bands, I have some follow up questions that I'd appreciate your take on (and anybody else that sees this post).

Basically, by analyzing the channel coherent bandwith and channel coherent time, as well as a few other things, I'm able to theoretically justify some choices of different frequencies and whatnot. I'm able to do some math, and even make some simulations on Matlab to justify my choices, and it all looks good on paper.
However, I have pretty much zero real world experience working with wireless systems. I want to talk with people who have actual experience with wireless systems and know what works/what doesn't.

The communication system my team is designing will likely be non-mobile for now, attached to sensors. It's also looking like the deployment area will be more industrial environments (things like oil and gas. So probably mostly open areas, but unreliable line of sight, and perhaps in the vicinity of large moving machinery, etc.).
So in such environments, what are the real challenges with wireless communications? What types of interference is generated in these environments? Is the interference narrowband, wideband, long lasting, short lasting? Will using 2 frequency bands actually help? What else can help?

Actually, after typing that out, I realize I should really get in touch with some guys with industrial wireless experience (maybe some engineers that have experience with wirelessHART?). In any case, one step at a time. My current task is to just evaluate the diversity aspect of the system. And getting advice and insights from many different people and perspectives is going to be helpful for me.