Contention ratios

In light of the broadband speed post, I am (again) going to try and explain why "contention ratio" is also a bad metric. Someone suggested it as a "quality" measure for an ISP, which seems sensible at first, but sadly not.

What is a contention ratio?

It is a measure of how much a link is shared (aka "contended"). The idea is you take the total demand on the link, and the link's capacity, and divide one by the other.

For example, if you have a 1Gb/s fibre going to a RAS which serves 100 separate 80Mb/s connections to customers: The demand possible is that all 100 of those customers may want 80Mb/s of traffic, so 8Gb/s total, but you only have a 1Gb/s connection which they are sharing, so 8:1 contention ratio.

The ratio is normally expressed as N:1, and a lower "N" is "good" because it means the link is not shared as much - simple... Basically, a link that is not shared so much is not as likely to slow down because of sharing of a link.

Sadly not so simple...

Contention ratio measured where exactly?

The first issue is you cannot really say the contention ratio "of the ISP" in a meaningful way. For a start, the end user is buying access to the internet, they do not really care if there is some arbitrary point that has been picked to quote contention - they want to know how likely it is that their communications slows down. But let's try an explain a little here using FTTC from AAISP.

For most end users the first contention is that there is likely to be more than one person in their house using the internet. That is contention. Their (share of) the link will slow down if the other people are also using it. However, we are trying to measure the ISP here, so one side of the ratio is normally the expected speed which is the sync speed of the modem the end user has. Sharing beyond that in to the home is not the ISPs problem. So that gives us one side of the ratio - add up the sync speed of all of the people sharing the "link" in question.

The first link is the customer modem linking to the modem in a cabinet - that is not shared at all - the phone line carrying your 80Mb/s (if you are close to cabinet) is not shared. So 1:1 contention, excellent - let's quote that shall we?

The next link is a fibre from the cabinet - that is shared with the hundreds of people on that cabinet - your neighbours. Openreach may publish metrics for that, I am not sure. We (AAISP) don't know what they are. A contention ratio measured there is not AAISP specific - but it is a perfectly valid concern that this could be the point your traffic slows down due to sharing. In practice Openreach operate VLANs and you can buy "lower contention" on a cabinet for more money.

The next link will be the cable link in the exchange. This is where a group of cabinets are linked to a switch and that links to a back-haul carrier in the exchange. We use TT and BTW. The sharing here is all of the people using that same back-haul carrier in that group of cabinets. I don't know if BTW or TT publish contention ratios, I doubt it. Again that is not an AAISP specific contention ratio, and will be different depending on which backhaul we have attached your service. Again, it is a perfectly valid concern that this could be the link that slows down traffic.

There is then an exchange backhaul where one or more cable links go to a metro node and other links before they reach a node that hands over to AAISP. This is, again, down to the back-haul carrier.

We then have links to carriers (BTW and TT). But these are multiple links and multiple sites, you will be sharing with all of the people on the link you happen to use, and that will change dynamically. So even if we worked out exactly what the total capacity of all of our customers is on each link, it changes if you (or they) reconnect. This is something we probably could work out if we tried.

We have links between switches, and we have links to LNSs, again, dynamic.

Then we have links to edge routers, and that will depend on where on the internet you (and other people) are trying to connect as to how much that is shared. These then connect to transit and peering, and again there is a contention.

Then out on the internet via transit, around the world there are shared links.

Ultimately you could say "what is the contention ratio" of a web site, e.g. aa.net.uk. That has a 1Gb/s port - so what is the total bandwidth of every internet connection in the world that may wish to access that site? Well the contention ratio there is millions to one.

Every single one of these shared points could be a cause of congestion, of slow down, so which contention ratio would you publish, and how would you handle that they can change dynamically. Even contention on your FTTC cabinet changes as other people take service in your neighbourhood.

There is no sensible way to come up with one meaningful point to publish as "the ISPs contention ratio". And even if you did, an ISP may have to publish several values for different services and back-haul carriers or even parts of the country. It would not be a simple number with which to compare two ISPs.

10:1 is not the same as 100:10 or 1000:100, really!

Even if you wanted to look at a specific link, and say what contention is that, it is not so simple. If you had 10 people with 10Mb/s links using one 10Mb/s shared link, then that is 10:1. But the chance of slowdown is high as you only get 10Mb/s if none of the other 9 people are using their service at all!

But what if you have 1000 people with 10Mb/s links sharing a 1Gb/s link, that is still 10:1. Now the chance of slow down is low, even if 99 other people are running their connection flat out at 10Mb/s you can still do the same. In practice, the larger the pipes, the more these things average out and you don't see congestion.

But both of these are 10:1 ratio, even though very different risk of slow down due to sharing. The ratio alone is not a good indicator of likely congestion - you need to know the size of the links as well.

Is the link congested?

The issue is actually congestion. Is a link slowing down because of sharing, basically is a link getting full.

The contention ratio does not actually tell you if you will get congestion. It depends on the usage of the shared link - how much do others sharing that link actually try to use - so an ISP that does a lot of "telemetry" type of customers (bus stop signage, etc) may have a massive contention ratio (if you worked out where to measure that) because their customers all send a few bytes a second, but they may see no congestion. An ISP that targets customers doing 24/7 UHD streaming would need more capacity per customer (a lower contention ratio), but even a low contention ratio may leave people buffering because of congestion.

The ratio is only meaningful if you know the usage level / demand as well. So basically, it is not meaningful.

Indeed, as one half of the equation is the expected speed, an ISP only selling 2Mb/s links would have a really good contention ratio because the backhaul links in the networks are all much larger now (to allow for all these 80Mb/s links). That would make such an ISP look good, when all it actually means is they sell slow services. Of course, if you have to advertise a contention ration, it makes sense to have an FTTC service capped at 2Mb/s, and quote that contention ratio as your headline package.

A better metric would be the average throughput per customer on a link. This is not a ratio, but Mb/s, and means the speed of the end links is not a factor any more on the assumption that people will use what they use (on average) regardless of their line speed. You still have the problem of which point in the network you are measuring though, and the problem of it mattering what actual usage/demand is.

What do AAISP do exactly?

At AAISP we aim not to be the bottleneck (i.e. not have the links we control getting full). All links are shared, so it is always possible, but we aim to invest as necessary to allow for changing trends in usage. We also challenge the back-haul carriers if we see congestion in their network (which means monitoring every line all the time). This does happen, and is normally either temporary (waiting for someone to upgrade a link) or a fault which we get fixed.

Even more confusion when picking a broadband provider

It looks like the rules for the way broadband (internet access) is advertised will be changing, again!

It is confusing enough, and I'll try and explain a bit here, but I wondered what the root of the problem really is, and whether the proposed changes will do anything to fix it, or, as I suspect, just cause more confusion...

What is changing?

The rules are being consulted on via ASA and CAP (here). The main proposal is something along the lines of showing the median (a sort of average) speed that people can get for a package. There are a lot of complications about what this means exactly and that is part of the consultancy. There are some alternative ideas like ranges of speeds. This is mainly for one-to-many adverts rather than individual forecasts for a specific customer address.

The upshot will be that instead of seeing "up to 80Mb/s" or "up to 77Mb/s" you'll see more like "average 60Mb/s" or "average 50Mb/s" even though we are talking about exactly the same service as now. The numbers you see are likely to be much more "all over the place" and vary per ISP.

You have to wonder why though... What is the issue?

The main issue I see is that people feel that they should be able to make some sort of informed decision about the service they buy. They understand price, but most people understand very little of the nitty gritty detail of what makes one ISP better, or worse, or just different from another ISP. People want a metric of goodness by which they can pick an ISP, and speed is seen as that metric.

Is speed a good metric for deciding an ISP?

Well, maybe. The problem is speed is a very complex metric. I go in to more detail at the end. Even in its simplest terms where faster is better that only applies up to the speed you might need. If the most intensive thing you do is live stream at HD TV, you do not need more than 10Mb/s. If you watch UHD, you may need a bit more. If several people in a house all watch UHD at once, a lot more. But for most people, once you are getting to speeds like 20Mb/s, really, for most things, faster is not a lot of help. There are exceptions, of course. Some things need downloads of large files (updated to devices, games, etc) rather than just "live streaming" like TV and the speed relates to how long that takes. As I try to explain later, the ISPs "speed" may not actually help with that though.

But yes, speed matters when it is low - if we are talking say 5Mb/s, or 1Mb/s, then you are going to find you are limited in things you can do. At low speeds there is no "live streaming video" for you and downloads can really take a long time. The problem is that these are never the "headline" speeds. Even basic ADSL2+ available almost everywhere can achieve over 20Mb/s of data throughput on a short line. The issue is not all lines are short, and so you may get lower speed. But that is something you find out with an individual forecast for your address/line, and not based on a one-to-many advert.

This means, for most people, the headline speed is not actually that much help as all headline speeds are generally more than most people need, and they may as well decide on price in most cases.

Even so, people will want "the best" - I would buy 80Mb/s not 40Mb/s, even though I have very little use for it. In fact, in my case, upload speed may matter more - uploading photos and videos. Upload speed is often not mentioned even.

Are there better metrics?

Yes, not that people are likely to understand, and that is the problem... In practice, one of the key factors in deciding which ISP, is deciding which technology is deployed by that ISPs. For a lot of people the choice will be between the many ISPs that can provide internet access over their phone line. Then, there is the choice of that being via a link to the cabinet (closer, faster) or to the exchange (slower). Sometimes there is a choice of cable (coax) or radio, or even satellite technology. That choice of technology matters. One of the biggest factors in the speed of your connection is that link on the phone line, or "last mile", and that will normally be exactly the same regardless of ISP if using the same technology. So in that way, the old "up to 8Mb/s" for ADSL1, "up to 24Mb/s" for ADSL2+, and "up to 80Mb/s" for FTTC were perfect. Everyone said the same speeds so the decision was slow, medium, or fast technology - people were picking technology without realising it.

There are many other differences, and mostly these come down to an idea of "quality". How congested is the network the ISP buys in, or runs, for back-haul? How congested is the ISP? How good are the ISPs connections to other ISPs and transit providers? What transit providers do they use? Sadly these are not something that lend themselves to a simple comparable number that an end user can see as "this ISP is better quality than that ISP". Hence speed being the thing, the only thing, the customers typically look at (apart from price).

Why is speed a bad idea?

A key problem with "headline" speed is that people really do not understand "up to" in this context. If you buy a car that says it can do "up to 100mph", you understand that when in a queue of traffic ahead you cannot do that, but you expect that, in the right conditions, you could do 100mph. Broadband feels like that, some times you download something slowly. You may stream video at a few Mb/s. But you assume that when you want to download something big you will get the top "up to" speed quoted. That is wrong, the "up to" is that some people get that higher sync speed and some people get lower sync speed, and what you get is what you get (baring improvements to wiring and fixing faults). People really do not understand that, and politicians have actually said they thing it is outrageous that 90% of people cannot get the "up to" speed advertised (after rules changed to not actually be the top speed, but 90th percentile). Personally I think it is crazy that a single person could get more than the "up to" speed advertised. The whole meaning of "up to" is "not more than". It is like people horrified that 50% of schools have a below average (median) performance rating. 100% will be at or below the "top speed" you can get and that is not outrageous, it is what "up to" means!

So people feel aggrieved that they do not get "the speed that was advertised", and no amount of "up to" or "not more than" will help with that. The change to 90th percentile meant pissing of only 90% of such people. A change to "median" will mean only pissing off only 50% of such people. It does not address the underlying problems. It still means a lot of pissed off people and politicians.

What is likely to happen now?

If the rules change to median, or something similar (per ISP) we will see more and more variation in the numbers quoted. When it was maximum the technology could manage, the same figure applied to all ISPs. When it depends on the ISP deciding or measuring a percentile it will be different per ISP.

One ISP has already started refusing to install lines that are going to be too slow (maybe they offer a "different package" for such people, or maybe not). The result is their median and 90th percentile speeds are higher as a result, and people think they are better. In truth the actual sync speed someone will get is not better on such ISPs than anyone else using the same technology.

It would be easy for us to have a "high speed internet package" that is only available on FTTC lines where you can get at least 70Mb/s, offering a different "standard internet package" to anyone that cannot get 70Mb/s. I can then advertise this "high speed internet package" as "average 79.82Mb/s"* when everyone else selling FTTC is saying "average 50Mb/s" or something. Buying from us (if you can get at least 70Mb/s) would get the exact same sync speed as anyone else as the modem is in BT even - but my advert would look "better"... We actually try to be more honest, and hence will lose out to such people.

* I checked, average sync speed of all lines we have ≥70Mb/s is 79.827Mb/s!

What can I suggest?

Ideally we need some clear way for people to be able to make that technology choice when one exists. That would be a good start, and ironically the old up to speeds (max sync speed possible) would do that as everyone would say the same figure for the same technology. That may be all we need, or if something else to then separate ISPs: a quality indicator by some independent metric. We also need ISPs to be honest on policies like traffic shaping, and CGNAT and IPv6 and so on, for the tiny percentage of people that understand those issues.

Obviously the technology indicator would be better if not actually stating a speed. A bit like 2G and 3G and 4G mobile. People have no clue what those terms actually mean, they just know 3 is better than 2, and 4 is better than 3. Ironically if we used such terms (ADSL1, as 1G, ADSL2+ as 2G, VDSL as 3G), etc, it would help, but new technology sometimes slots in the middle and breaks simple ideas like that. If we had industry standard agreed terms for such, that might help.

It would also help if terms like "fibre" were not misused. Actual fibre has speed and reliability implications where even a small amount of coax or phone wire can have issues. It seems the ASA may be turning the corner on that one.

Why is "speed" complicated?

This is where I get in to a bit more detail and try and explain some of the issues with quoting speed. Some of these are touched on in the consultation, but most are totally ignored.

The end user will see a "speed issue" in two main ways. One is if they are doing something like streamed TV and they get "buffering", where it pauses to catch up, or worse it refuses to play "live" but has to download first. Some services like Sky TV are quite good as the default is to download, but even then, if you have a fast link it will let you start watching right away, and if not then you have to wait for some download. The other main way people see issues is when downloading a file of some sort - e.g. some device upgrade, or game upgrade or something you have bought and are downloading, like an app. Some game upgrades are gigabytes in size and can take hours.

The buffering issue is a simple binary one in most cases, if your service is fast enough you don't see it, else you do. It depends on content, and actually services like netflix are clever, they do lower quality if you don't have the speed. So the issue is seen as lower resolution rather than pauses and buffering.

But in all cases the reason something may be slower than you hoped comes down to a lot of different factors. Just some of which are :-

• Your device or PC may be old and slow. That is actually becoming an issue as PCs from only a few years ago simply did not expect to transfer data at the speed of modern lines. It will not be long before people have gigabit fibre links to their home and still have 100Mb/s connected machines, or 50Mb/s wifi, and a machine that cannot even manage that speed. We can't all afford the latest and fastest machines.
• Sharing at your home/office - where other people are using some of that capacity, even when you do not realises. This is especially confusing when devices download upgrades by themselves and you, alone in the house, find things going slow when in fact the ISP is delivering a good service to your home.
• You are using wifi - this is a big issue, and confused by the fact that wifi standards exist that even go over 1Gb/s. The exact type of wifi in your access point(s) and devices will impact the speed, and internet connections can be faster than your wifi could do, even in ideal conditions (especially if old equipment). This is compounded be people having a single wifi access point under the stairs that is part of the free router they got. A house of any decent size, or if it happens to be Welsh or Cornish (OK, not that fair, but they do seem to have a lot of places with stone walls!) will need more access points for good wifi coverage.
• The modem itself - newer modems are better and get better speeds, but this is less of an issue these days as the chipsets have come a long way.
• Protocol overhead - a tricky one as the protocol overhead is part of what is sent on the wire, so why not included in the speed. But people expect something that matches what their screen shows in terms of bytes per second, so people expect the protocol overhead to be removed when quoting a speed. Usually it can be worked out as a fixed percentage.
• The line (coax or telephone wire) will be one of the biggest bottlenecks. Even when a fixed speed link on fibre, it is likely to be the bottleneck compared to your PC and the internet. When not fibre, it is one of the most variable issues between customers and hence the main reason for "up to" speeds. Other than faults, or bad internal wiring which can be fixed, you are mostly stuck with what your modem can achieve using the specific technology. VDSL (to cabinet) is better than ADSL (to exchange) where you have a choice (not actually always the case on some long lines). Fibre is better still.
• Once we get passed the modem links, we are in to things which are not customer specific, but can be regional, and can be ISP specific - like congestion to your cabinet. Now, BT (Openreach) are pretty good on this, and ISPs can pay for better services (lower contention). We generally do not see this being an issue, yet.
• Congestion on back-hail network - linking all these people to the ISP. This may be part of the ISP network or totally separate as a wholesale access network. This can have issues. Again, using BT and TT backhaul, any congestion we see is almost always temporary (waiting to upgrade a link) or a fault (which we chase to get fixed). Even so, the back-haul do not actually sell an un-congested service to us.
• Congestion in to ISP - this is perhaps more of an issue as typically the link from back-haul to ISP costs by the capacity of that link. This means ISPs may well dimension their links so that at peak times the link is full, and then take steps to make it a fair way to access the internet without some people hogging all of that capacity. This makes for complex shaping systems.
• Shaping systems within ISP to manage capacity and keep costs down.
• Capacity within the ISP themselves - not so hard to get right.
• Links to other ISPs and transit - this is usually less of an issue as such links are much lower cost.
• Now we get to things that are well outside the ISP control, such as congestion on peering and transit links. We have seen on peering links, but not generally on transit links. Of course it is ISPs choice which links they have and how big they are.
• The next ISP - the ISP of the service you are actually trying to access - they have all the same congestion, capacity, and shaping issues.
• The end servers themselves, policy based speed controls. We have seen this on downloads where clearly a service has capacity, two downloads at once are fine, but all downloads are limited to, say, 10Mb/s.
• The end servers themselves just over capacity - can happen and things get really slow. This can be noticeable in some game upgrades. It depends how it is done. There are content delivery networks which are really good at getting this right.
• And finally, faults. The way faults are detected and managed is crucial. Lines can be slow because of a fault that needs fixing.

So, when someone sees something being slow, the actual slowness could be any or all of the above. Some are end users fault, and can be fixed (use ethernet not wifi, fix internal wiring, etc). Some are the backhaul provider (contracted by the ISP), and some are directly the ISPs "fault", but then a whole load are not the ISPs fault. They are "the Internet" which is a shared service at heart.

So, naturally, if an ISP quoting speed in an advert is trying to allow for not just sync speed but also some aspect of "peak time slow down" where do they draw the line? As an ISP I would say it is speed in our own network, and for that we, as an ISP, we (AAISP) don't have a "peak time slowdown" at all.

You then have issues of what happens when an ISP has a speed tester and deliberately prioritises traffic linking to that speed test...

See, it is tricky!

P.S. My son's view on this...

Don't allow one-to-many adverts to state a speed at all. They have to sell on something else. It becomes the razer blade adverts, 5 blades better than 4, obviously.