4G (HSPA+) and Cable Broadband in Ewelme
The previous penning of pages about ways of obtaining broadband in the rural village of Ewelme, was prompted a few years ago as an attempt to convey to the local 'broadband have nots' or 'slow have gots', that notwithstanding rather gloomy news about Ewelme being a 'not spot', there was/is indeed perfectly sound and often faster alternatives to that provided locally by BT owned land lines. It is now known that during that period, some 40 households in Ewelme (how that equates to the actual number of 'users' is unknown), have adopted the use of wireless/mobile broadband (3G+) for themselves, and it would not be unreasonable to assume that others who have not needed advice or required assistance, are also now utilizing this technology within the village boundaries or in the near locality (e.g. RAF Benson, Berrick Salome/Roke?). Although the information provided previously is still highly relevant, we are now moving into a new era, where on the one hand, wireless broadband has evolved into an order of magnitude faster than landline performance here, but where news of the forthcoming effects of government funding is also starting to emerge, in terms of the attempts to alleviate the gap between broadband speeds available within rural areas and those experienced within more densely populated towns and cities.
This page (last update 20/2/12) is therefore intended to describe the current and still evolving 'state of the art' of mobile/wireless broadband, but in view of expectations of the 'soon' to be available fibre optic cable to the central cabinet in Ewelme, discuss whether this means that our village is to become a centre of multi-technology communications excellence, and how one can compare the two technologies. One could confidently predict that it can only be a matter of time before the equality of available broadband speeds throughout the land will be achieved, in the same way that we all now have access to electricity and telephones. However, this example also needs to be tempered somewhat by the fact that nobody in Ewelme can yet obtain mains gas! Indeed, when this page was perhaps rather prematurely written in November 2011, exciting local news being conveyed via the Parish Council indicated that both Wallingford Exchange and Ewelme itself would be enabled using fibre optic cable for 'Super Fast Broadband' to be installed by the end of December 2011. However, on the 15th February 2012 the front page of the Wallingford Herald confirmed that previously BT rescheduled upgrades to Wallingford Exchange to enable 'Fibre To The Cabinet' (FTTC) installations would indeed take place from June 2012, and Oxford County Council are proposing that rural areas associated with this exchange will be regarded as a BDUK funded pilot scheme prior to other rural areas in Oxfordshire getting the same service by 2015. Although it is yet to be established as to what residents of Ewelme can expect from a fibre optic cable service some 5km from the exchange, the following information may be useful not only to those already using Wireless Broadband locally, but also as a true means of comparison when cable eventually comes to town.
E586 HSPA+ MiFi router and cradle.
Wireless Broadband Today
Now although this page is titled '4G' to differentiate it
from the previous '3G' descriptions, we haven't yet
officially reached this particular stage of propagation of
broadband by wireless (also known as LTE - Long Term Evolution). However, it has
recently become apparent that this local area at least, is
now capable of servicing mobile phones and wireless broadband
dongles and routers designed to operate using what is
technically known as an HSPA+ network (High
Speed Packet Access
plus). To put
this as simply as possible, 3G and its evolutionary upgrades
were/are capable of downloading data via wireless at initially
maximum speeds of 3.6 Mbs, then 7.2 Mbs. Firstly, it required
that not only the 'Three' radio masts on the Sinodun Hills (Whittenham
Clumps) or at Hale Farm Benson needed to be regularly 'upgraded', but
so also the
performance of marketed dongles and routers! So, if you are still
using equipment bought when 3G first became available, you will
only be able to receive data at a maximum speed of 3.6 Mbs.
Similarly, if you are using the still available E585 MiFi
router/dongle, the maximum speed you will be able to experience
will be 7.2 Mbs (the highest recorded speed published in Ewelme
using that device was 6.8 Mbs).
Local browser display of HSPA+ E586 MiFi router showing high speed capability.
An HSPA+ network enables data rates to operate at up to 21Mbs, although it is understood that the local mast and receiving equipment capability is currently publicized as being 40% higher than it was before, which implies a 10 Mbs maximum at the moment. However, using the older equipment you will not be able to notice the available speed increase. The highest instantaneous peak speed recorded in Ewelme to date is 8.521 Mbs using the newest E586 MiFi router/dongle, and generally speaking, daily operational speeds are indeed higher than those experienced before the hardware upgrade occurred. However, it is difficult to quantify whether this always equates to 40%. Upload speeds have also noticeably improved, and 1.73 Mbs has so far been locally recorded.
See also http://blog.three.co.uk/2012/02/01/introducing-broadband-in-a-box/ for Three's new HSPA+ 'Cube' device and http://www.ruralbroadband.co.uk/3g-wibe-by-deltenna-with-rural-broadband-firmware.html for a device that practically guarantees at least 2Mbs wireless broadband at a higher price, if you are not convinced that you can receive broadband in a rural area.
So, this page is not intended to repeat previous advice about the mounting of dongles/MiFi units on windows, neither to provide information on the different providers that are available, or compare equipment. Over a long analysis period, it has now been concluded that the best solution for obtaining broadband in Ewelme is by taking a monthly or longer contract with the company 'Three' and by using the latest MiFi dongle/WiFi router E586. What follows is a discussion on how to test your broadband performance, and when fibre optic cable becomes available locally, how to quantify and to compare it with what wireless broadband has to offer. Superfast Broadband is regarded as offering speeds of 25 Mbs. LTE (4G) trials currently underway in the mountains of Wales are apparently testing speeds of 50 Mbs, using some of the newly available frequencies left following closure of analogue TV. Exciting times eh?
Is Speed Important?
Well, it all depends on how and why you want to use your internet
connection. If you want to use it just for emails, browsing
social interaction sites (e.g. Facebook/Twitter/Messenger) and for doing some internet shopping, I guess your
main concern will be whether the response you get from a mouse
click or keyboard 'Enter' does not include noticeable delays or
slow screen updates. It is well known that if you use 'dial
up', such usage can be painfully slow (56 kbs max), and
hopefully anybody who had such a connection has now at least got
hold of the lowest 3G offering. On the other hand, if you have
Ewelme type landline broadband (0.5 to around 1 Mbs variable), the same
usage regime might well be deemed acceptable, possibly depending
on which side of the village you live. However, if you want to
experience anything that involves the transmission of
video, such as watching a missed TV programme or browsing
YouTube, downloading a
large file containing a film or software update, holding video conversations or
playing interactive on-line games, you need to know that the
current and variability of the particular network to which you
are connected, is robust enough for your usage requirements.
You can certainly use Ewelme landline broadband to 'stream' (play while downloading) an internet radio program, which requires less than about 0.2 Mbs continuous bandwidth (speed requirement or availability). But when it comes to watching a Standard Definition (SD) video program from the BBC or ITV, you need a bandwidth of at least 2 Mbs to prevent the performance constantly stopping while the data download 'catches up' (buffers) with what has already been displayed at a faster rate. It is unlikely that you can do this using Ewelme landline broadband without constant buffer delays. However, depending on your mobile broadband equipment age and status, you should be able to watch such a programme without interruption during extended periods of the day. When it comes to watching High Definition video, the bandwidth requirement rises to a minimum of 3.5 Mbs for uninterrupted viewing. This would only be possible at certain times during the day using a 7.2 bandwidth 3G router/dongle, but with a 40% increase in possible efficacy using the new 21Mbs dongle/routers, ensures an even greater likelihood of success.
Do you really want to watch TV using the Internet? Well of course you do, because the facility is there! But apart from it being a useful occasional feature if you have missed a favourite programme or there is nothing else to watch on the countless digital channels now available, it should be considered that recording and replaying your own digital programme using FreeView, FreeSat or Sky is so easy now, and doesn't use up any of your monthly download allowance, or use up the available bandwidth of others. I'm sure these comments aren't going to put you off still wanting to have the choice of having a go when you want to, and I would also share those sentiments.
How is Broadband
Speed Measured?
The classic method of determining internet speed, is to
communicate with a dedicated server (remote computer), that has
a high speed internet connection. A file is then downloaded to
your computer, where knowing the size of said file and measuring
the time it takes to complete the transaction, determines how
fast the connection is operating. The opposite is then performed
to determine the upload speed. It can be categorically stated,
having had many years experience in testing communication
speeds, that you will never get the same answer twice! However,
such a test methodology does give a good idea of the usability
of the Internet at a particular point in time, and by recording
many readings over an extended period, a fair idea about the day
to day performance of your local network can be established. Why
can't you get a consistent answer? It's because an internet
connection between your PC and the file server consists of a
number of intervening routers and physical communication
technologies. Take for example performing such a test by using
local wireless/mobile broadband. First, assuming the use of a
MiFi router, the initial connection is between your PC adaptor
and the MiFi via WiFi. N.B. It is just as important to ensure
that your internal WiFi signal is a strong one, and not one
struggling to tunnel itself through walls and doors, since this
will have just as bad an effect as a weak external 3G/4G signal. This is then physically switched (routed)
via an HSPA+ connection to your local radio mast. From there it is
likely to be transported (backhauled) via fibre optic cable to a
high speed internet 'backbone' (router). Who knows what routes
it will then take to get to the speed-test server, but
most likely to be via
several routers connected by fibre optics or microwave links. If the
server is in another country the route may also involve
satellite.
Data is transported in 'packets', which means that a file
is sliced up into chunks of, for example, 100 bytes. Each packet
contains a destination address (Internet Packet - IP address)
and an origin IP address. The physical route each packet takes is
determined by each router on the journey, and may be different
for each packet. It may be that one packet arrives at its
destination earlier than
its predecessor! However, each packet is numbered and receipt of each is
acknowledged by sending a short message back to the sender (one
reason why you also want a fast 'upload' speed), and
the complete file is then reassembled into it's original form by
the receiver. Packets that get 'lost' and are not acknowledged
for a variety of reasons, are resent. Such situations obviously
slow the whole process down. Now consider that many other
packets are also being transmitted and received from
various unrelated sources and destinations within each of the
sections of the complete route, just like letters in a postal
system at various sorting offices. The more packet traffic there is, the slower the system
speed for your packets will be. Also, each section along the
route may have a different 'bandwidth' which is a measure of its
ability to send packets at a certain speed (i.e. in Mbs). So you
can see that for a given speed test, it will certainly indicate
the current trend, but is unlikely to give precisely the same
result every time. This description also applies whether the local
broadband connection is one of fibre optic cable, telephone wire
or wireless, using the communications technique known as TCP/IP
(Google it for more info!). Of course, all this happens so fast, that as users,
we don't notice that we are sending and receiving millions of
packets of information!
This dynamic illustration tries to convey the multi-route journey individual packets might take in transferring pieces of a file from one remote computer to another (Your PC at the bottom) . What is not shown are the unrelated and random packets travelling between other sources and destinations that would also be included during this file transfer. This would have the effect of determining the instantaneous speed such a file transfer will take from moment to moment.
You can see for yourself how many 'hops' there are between you and, for example, a web server you are browsing. Open the 'Command Prompt' available on a Windows based PC, and use the 'tracert' command together with the URL (named address) of a web site, as shown. This will return the number and identity of all the routers a single test 'packet' has had to go through to reach the destination.
Speed Test Methods
There are many speed test facilities available,
basically using the principle of downloading a fixed size file and
measuring the time it takes, then returning the file back to
establish an upload time. One of my favourites has always been
mybroadbandspeed.co.uk, which presents an uncomplicated and
advert free tester, but still with the facility to record a
limited number of test runs, and will even draw a graph of
historical performance. However, my functional preference
at the moment is the
BBmax tester, which although a bit garish, provides some
very useful logging facilities. There's even one,
Speedtest.net, that lets
you choose from test servers located all over the world, but
that might be going a bit far don't you think!?
The first thing that precedes the BBmax speed test proper is what is known as a 'ping' test. As its name implies, it sends out a very short message to the IP address of the test server, and measures how long this takes to be 'reflected' back. Such a test not only gives some idea of how clear the route to the file server is at a certain point in time, but also an indication of the reaction time (latency) of the network, which could be important to someone playing an on-line interactive game or video conferencing. The BBmax tester also allows you to keep an unlimited on-line log of your results. This has proved to be most useful, not only to see long term trends in speed results, but also allows the test to be both portable in terms of being able to be used on different computers and smart devices, but also in different locations. To do this, you need to log on after you have registered a user name (email address) and password. This is exactly what 'Three' will ask you to do if you are discussing some sort of performance issue with them. They usually want you to do some tests over three days! You provide them with your email address and password, and they can see what's been going on. Would you like to try mine to see many months performance? When you reach the site, log on as chris@jenkins-of-ewelme.org.uk with a password of '333333' and look at 'Previous Test Results' to see the increasing broadband trends over the past 18 months. In recent times, there have been some, what were thought to be 'ridiculous', results e.g. >15 Mbs. But because these sorts of recorded speeds are happening more frequently on a regular basis, they just have to be believed. Anyway, would you be so kind as to remember to log off again and then perhaps set up your own log? Thanks!
As
an example of portability, here is a snippet of my log, where
morning readings had been taken using the home based MiFi, but
in the afternoon I visited Benson Library (Oxfordshire fibre
optic cable network, also used by Ewelme School), and confirmed
the slower speeds regularly measured here. On return home soon
after, the speed generally doubled via wireless broadband. Such
fibre speeds I'm sure would be expected to significantly
increase after the expected investment in the infrastructure?
It's all very well doing these spot checks occasionally to see whether you are getting satisfactory performance, but my main means of continuous 'real-time' monitoring is to use the 'tbbmeter', which you can download from http://www.thinkbroadband.com/tbbmeter.html . This can be compacted to be a small inconspicuous square and arranged to be 'on top' of your display so you can 'see' all the Internet traffic that comes and goes through your PC or device.
This
display is showing the varying speeds of buffered data download
while watching a TV programme in High Definition. Note the
instantaneous speed shown of 8.5Mbs!
Other forms of Test
The tests above do provide an idea of instant broadband
speeds from moment to moment throughout the day, but what about
comparative quality? Here's one which recognises the
difference between a test using mobile/wireless broadband from
this address and that currently possible using landlines from
Wallingford exchange.
There is no doubt about the potential superiority of wireless broadband at the moment. But what happens when cable is installed in the village? This is completely unquantifiable until physical tests can be performed. However, the results of national average speeds give an indication of perhaps what to expect. Do we assume that we will get the sort of speeds offered by Virgin Media or those offered by BT Broadband? Try it for yourself?
Now, here is a tough test run provided by the BBC, but one which will provide a good indication of the 'quality' of your broadband speed. Rather than just downloading a known sized file and measuring the time to arrive at the linear speed, this one simulates the actual buffering protocol used when watching or listening to a streamed TV or radio programmed. It also does it three times and then takes the average, and provides a realistic idea of whether you can just listen to a simple radio stream or have the bandwidth to watch a full high definition TV programme.
You can certainly see the difference between an 'ordinary' speed test and the actual data download while streaming. At the time of writing, users of landlines in Ewelme would be hard pressed to achieve much more than streaming radio, and to watch HDTV would be out of the question. Although this test doesn't always produce this result, it does indicate that it is likely that standard definition TV can be streamed at any time using wireless broadband in Ewelme. Try it for yourself?
Is Fibre Optic
Cable Broadband Coming?
For many years, the news in Ewelme has been that the
likelihood of obtaining an improved land based broadband system
using fibre optics cabling would be some time away, due to it's
weak commercial feasibility for companies such as BT or Virgin.
However, due to various national campaigns and government
lobbying, there is now a real possibility that rural areas such
as our own may be eligible for what would need to be a significant
financial investment. The coalition government has set aside
£530m for distribution across the UK, to be used by regional
authorities for the purposes of ensuring that areas not serviced
with usable broadband (regarded as 2 Mbs minimum) will be
provided for. On 16 August
2011, BDUK (Broadband Delivery UK) announced that Oxfordshire would be
allocated £3.86 million to provide faster broadband speeds in
rural areas over the period to 2015. OCC will be able to draw
down this funding on completion of an agreed
Local Broadband Plan. However, OCC recognises that the
funding will be insufficient to achieve complete minimum speed
coverage in all rural areas by wireless and fibre, without the
addition of private and perhaps local funding. They will be considering
proposals during 2012 for implementation during 2013.
However, the County Council is to launch the Oxfordshire
Community Broadband Fund in November to which it is hoped that
communities will bid into for small community projects.
This would indeed be good news for the village, for it would
mean that those currently using landline broadband who have not
been inclined to try alternatives, could soon
upgrade to a much faster means of Internet connection. (Is there
anyone still on dial-up?!) What speed might this service offer?
This is completely unknown at this point in time, and might not
be established until tests are performed at an upgraded
location. This was basically the same situation that occurred
following the installation of the 50 core cable in 2007, where
eventually it was discovered that the maximum speeds achievable
were between 0.5 to around 1 Mbs depending on the distance and
quality of the cable from the Kings Pool cabinet, which I guess
we were all grateful for for a while. What such an installation
would present, would be a choice between what is offered by mobile
and fixed wireless based broadband, and that provided by fixed
fibre optics, where Ewelme could actually become the envy of similar
rural communities within Oxfordshire for a while!
Comparing Wireless with Fibre Optic Cable Broadband
Apart from the way the broadband router box is
installed, where for cable, it will be connected to the incoming
telephone lines (BT Home Hub?), and for wireless, it would be accessible by WiFi
from an optimally sited (at/on a window) MiFi router, there would be
little to chose between the two, except perhaps for speed! My
only suggestion is that one of the first subscribers to cable
ought to be the
Ewelme Store, which currently
copes with landline broadband, but which could not only usefully speed up on-line
credit/debit card transactions, but also boost the speed of
the currently Free WiFi for visiting customers to
the cafe.
Here
is the remotely monitored Ewelme Store router real time on-line
'statistics' display. This shows the current speed potential of
0.7 Mbs download and 0.45 Mbs upload. Hardly broadband right
now, but if connected to fibre optic cable via the BT
cabinet a few metres away, would indicate the best speed
potential by cable in Ewelme.
Having the ability to freely connect to such a
service would mean that anyone could test the speed of Ewelme's cable
broadband using a WiFi enabled PC or Smartphone
(there are many speed test 'apps' available as well as those already described), from a point only a few metres from the central FTTC (Fibre To
The Cabinet) while sipping a cup of coffee! The Ewelme Store
router is also currently being monitored remotely, so probably
the performance of the highest quality cable access point in the
village can act as a regular comparison with wireless/mobile
broadband. However, it is well known by now, that the indigenous
aluminium telephone cables can drastically attenuate a high
speed data signal. In the case of current landline broadband the
effect has been shown to cut the speed down by 50%!
It is thought unlikely that a 25 Mbs superfast broadband
capability provided by Wallingford would also 'appear' at the
central box in Ewelme, due to the underground distance from the exchange
(some 5 miles?) and
shared usage,
although the attenuation over distance for fibre optic cable is
far less than that exhibited by the 2007 installed copper cable. However,
would it still be the case, that the outer reaches of the
village would still only expect half the speed seen at Ewelme
Store? Perhaps the Parish Council ought to finance a test at
'Lower End' (some distance from King's Pool), when cable broadband becomes available, to establish
the span of broadband speeds across Ewelme!! Will the process
continue of one technology alternatively leapfrogging the other,
before reaching such ridiculous speeds that no-one will care
anymore? In the meantime,
enjoy the 'up to' 9 Mbs broadband currently available in
Ewelme by
wireless!
Conclusion?
It can sometimes be difficult for someone trying to establish
what level of broadband service they might get, before
they sign up to what is often an obligatory 12 months. This page
and previous ones on the same subject, tries to provide a means
to 'try before you buy' so to speak, by demonstrating various
ways to test local broadband speeds against the advertised 'up
to' hyperbole.
There is little doubt that evolving data communication using cellular
wireless technology is here to stay, increasingly supported by
the proliferation of smart phones and iPad like tablets
incorporating 3/4G interfaces, or the increasing use of WiFi enabled laptops
connected to a local cellular mast via a fixed or portable 3/4G
router. Some smart phones now offer 'tethering', meaning they
can act as a 'dongle' when connected to a PC, or can
also operate as a WiFi access point (AP) to provide 3G access to
other local PCs, and some even
offer 'all you can eat' data usage for a month for a one
off payment. However, for these smart phones to act successfully
in this area, would mean that they would require a 'Three'
compatible SIM card to be installed. The question is whether cable and wireless performance
will remain comparable in terms of speed and service. The
advantage of wireless technology over cable is not only that it
is capable of dealing with home based voice, email, web and HD
TV, but it is also portable, enabling it to function during
travel, at work and on holiday. The advantage of cable over
wireless is that if it can be provided in areas such as ours at
a reasonable speed, then consideration of distance to and
direction of a local radio mast, your terrain or the weather, will no longer
be such an issue. Maybe in
time though, a multi-provider 4G mast will be inconspicuously situated at the top of Green
Lane or high up behind the Ewelme Rectory? The history and
evolution of some form of rural broadband in Ewelme now spans
7
years! (see
Ewelme Wireless Broadband and
3G in Ewelme), and has shown that wireless technology doesn't stop still for long. So, it
will be most interesting to see what the next couple of years
will bring us in this regard. Isn't technology wonderful,
or what do you think?