Hi Experts, anyone familiar with backhauls?
Does 10 Gbps capacity fiber produce = 10 Gbps capacity microwave in terms of average user throughput experience?
I know backhaul affects user throughput when it’s limited but here they have same capacity.
Please elaborate theoretically and practically and what would be good KPIs/constains to judge here.
Are you using fibre or microwave?
And for what purpose are you using this backhaul like DSLAM, 3G, 4G or 5G, and standalone site or is it going to be shared with clusters of sites.
Generally speaking, bandwidth of fibre is infinite but the limit is transmission equipment which is currently limited to some 160Gbps. While configuring backhaul, it is very important to check the end to end port capacity. Eg if we have a 1G port at one and and 10G port at another end assuming auto negotiation is enabled at both end, the throughput will be limited to 1Gbps only.
But as microwave is a wireless radio based technology, there are limitations on the use of Spectrum use of the Govt of the country so as to avoid interference with critical wireless services. And only limited licensed bands are allowed to use as microwave links and hence the available bandwidth is very limited.
Earlier the backhaul capacity of Microwave was limited to 140 Mbps but now it is tested upto 100Gbps upto 1.5 km. I have worked on a microwave link of 100 Mbps capacity over a distance of 20 km. But now the technology has evolved to get good capacity over a longer distance. Line of sight is very important in microwave. And if we have sufficient frequency spectrum, and the link is healthy with good SNR, we can achieve good and reliable service in practical scenerio. The best thing about microwave is that it is maintenance free and very cheap in implementation.
As far as bandwidth is concerned, this can be tested using equipment like Ethernet tester between end to end links. We can test the bandwidth using a PC as well given the PC has the required capacity port ie 10G port using softwares like iperf, jperf, winscp, etc.
Both fibre and microwave is a symmetric ie capacity will be same in both direction ie uplink and downlink will be same. So there there should not be any difference in throughput or performance. If any deviation observed, then Ethernet tester can be used to check packet loss, jittering, etc.
I hope that helps.
Thanks a lot for sharing these details!
I actually want to know that aside from capacity, can there be differences in throughput only due the fact it’s a MW vs a fiber? For example is BER/packet loss is usually higher in the case of MW which can affect throughput on the end user?
Also a friend told me that the MW utilization doesn’t directly affect user throughput and I find it strange.
So if we have a NW that uses fiber mainly vs a network that uses a MW mainly while capacities and other factors are identical, can we expect the fiber NW would have a more stable NW and a higher avg throughput for its users?
No, If MW link is healthy, there should not be any difference in throughput.
Once link is locked with good SNR, we can configure it with required bandwidth.
If link margin is good, link will be stable.
SNR and link margin is very important.
How about the utilization part?
Do MW links get utilized more than fiber do?
How can utilization, modulation, congestion control differ between MW and fiber?
Given that the same amount of traffic is passing in both cases of course, I’m referring to BH capabilities only. For instance in a 5G NW which is currently an NSA but could potentially upgrade to SA. Also with the growing traffic trends. What would it take for MW links to match with fiber links?
I think if you can elaborate on specs as well as KPIs to look out for it will be a lot clearer.
At the end I want to benchmark between two deployment scenario, where I can judge that current BH is either sufficient and delivering best results to end users, or no, they should start using more fiber and/or higher specs MW
That’s why I asked initially about the requirements.
Commercially available microwave link is now limited to less than 6 Gbps.
Theoretically 5G throughput is 10 Gbps.
But remember this the maximum we are talking about.
This will vary wrt to spectrum bandwidth, MIMO size.
In such cases, fibre will be a preferred backhaul media.
But if you are finding it difficult to lay fibre where terrain is very difficult, still microwave can help, with low capacity 5G with speed limited to 1-2 Gbps.
We use EBand then one can upto 10GBPS as well.
Good read, thanks.
Yup, I’ve seen this deployed and I’m thinking practically in the field, is the availability of this capacity close to 100%?
I read that it there should be a supporting lower band MW to ensure longer distance transmission, how can this affect the actual capacity and therefore end user throughput.
Eband could be 1 km hop.
Majorly small cells mesh or fiber going to be sort after.
cbd still fiber can be planned.
So maybe eband can help with 5G mmWave deployment as it requires high number of small cells.
But in the case of macro sites where distances are higher it’s not a practical solution?
Also one important question how can high utilization affect user throughput?
In radio we know that prb utilization will directly impact user throughput, is it the same in the case of MW transmission?
It’s QAM, No of MHZ slots, MIMO or 1+1 increases throughput and resources.
No it’s not like that. MW is a directed beam end to end with line of sight.
You can expect more than 99% transmission efficiency.
If you configure 1 Gbps, you will definitely get close to 1 Gbps.
Mobile technology should not be compared with MW as microwave technology which uses directed beam in only one direction.
It is mostly used for backhaul.
Thanks for the input @shamshad and all, great insights!
I just managed to reply as I had a busy day. 
Just a final concern, any impact on transmission latency and/or packet loss for Fiber vs MW?
If link is healthy, latency will be a few ms and there won’t be any packet loss.
No jittering as well.
Hello,
Keep in mind MW is over the air and can be affected having fading by rain, obstacles, vegetation, etc. which will end up affecting modulation and ofc bandwidth. They are usually worse to troubleshoot and often require 2/3 technicians if you have to work in the tower.
Fiber on the contrary is easy to upgrade and test, and less prone to degradation, but takes longer to deploy to reach the endpoint.
My advise would be deploying by MW on complicated sites, while you take your time to deploy fiber to reach it.