SeeIT Consult » broadband

Imagine Wi-Max: Three Weeks On

admin — Tue, 15 Jun 2010 15:50:49 +0000

I have now been using the product for 3 weeks and have 21 days worth of upload, download and ping data. The latest graphs have been added to the end of the post: Imagine Wi-Max: One Week On.

Imagine Wi-Max: One Week On

admin — Tue, 25 May 2010 17:54:15 +0000

This is a follow up to my initial review of the Imagine Wi-Max product 2 weeks ago.

Updated on 2010-06-15 (see end of article for 3 weeks worth of test data).

My last post referred to some shortcomings with how the Wi-Max modem was shipped with an inaccessible control panel, making it all but useless to anyone but the casual home/non-technical/non-business user.

Since then I have:

received a replacement modem (hardware fault in the original)

obtained access to the control panel (allowing me to set up my network properly)

had about one week of general use

performed some bandwidth tests

Modem indoor reception

The modem has 4 lights indicating signal strength to the Imagine mast (high site). In Dublin 7 I was getting 4/4 lights lit up with the modem close to an exterior window. I guess that I am about 1km from the nearest Imagine Wi-Max mast, so I can rule out problems associated with being in an area of marginal signal coverage.

However the window location was too far to cable to my server, so I moved the modem. At the moment the signal strength lights are 2/4 and it is not even in line of sight of an exterior window. No problems to report.

The manual says that at least 2 lights should be lit to ensure reception.

Test methodology

Tests are conducted from a Linux server cabled directly to the Imagine modem. I am on a 3MB down, 512kB up tariff.

Ping test: 50 pings to ftp.heanet.ie

Download test: ftp download of a compressed (bz2) 4.5MB (approximately) file from ftp.heanet.ie

Upload test: ftp upload of a 1.2MB bz2 file to a virtual server in the UK

Tests are conducted every hour and the initial results are from a single 24 hour period. The telephone service was not tested at this time.

Important: There are about 5 or 6 clients on the network connected by Wi-Fi. Their usage is average to low, but the server is not guaranteed exclusive access to the broadband ‘line’ during these tests. However seeing as it is the only PC cabled to the modem and the tests run over a reasonable, it shouldn’t make too much of a difference. As time permits I’ll do some exclusive tests.

Responsiveness

Compared to the DSL connection I was used to, this definitely feels more sluggish. Ping times are longer than I was used to and are not as consistent, but overall it cannot be compared to 3G, Ripwave or dial-up.

The graph (click for a larger version) shows a record of ping times to a nearby server. The minimum ping time averaged at 59 milliseconds. Average was 135 ms but the responsiveness did vary quite a bit.

Given the variance in the response times, I am a bit curious about the performance of the telephone line. Voice over IP is particularly sensitive to these variances.

Hour Min Ave Max StdDev

0 76 166 1064 164

1 43 82 192 34

2 63 63 63 0

3 62 135 793 136

4 61 115 937 123

5 80 369 1644 342

6 61 93 221 36

7 46 90 262 36

8 32 65 103 10

9 76 83 128 16

10 74 149 1084 188

11 63 109 818 107

12 59 82 141 14

13 65 78 129 8

14 63 129 1307 176

15 60 103 525 79

16 46 82 158 16

17 31 83 194 33

18 54 81 143 19

19 51 86 197 28

20 40 168 759 140

21 76 258 1536 347

22 61 106 834 110

23 60 80 143 15

Speed

As measured by hourly upload and download. Again this was not too bad. I am on a 3MB tariff at the moment and the average download is 2.61MBps. Without knowing what the contention ratio is it is difficult to comment, but that performance is pretty reasonable for a home/small business user on a highly contented product.

Hour D/L U/L

0 2.23 0.49

1 1.73 0.39

2 3.10 0.48

3 1.80 0.46

4 2.99 0.43

5 0.99 0.40

6 2.09 0.43

7 4.08 0.44

8 3.44 0.48

9 2.51 0.41

10 3.52 0.46

11 3.99 0.48

12 4.78 0.49

13 4.15 0.49

14 3.62 0.47

15 2.59 0.44

16 3.76 0.48

17 3.29 0.47

18 3.75 0.46

19 2.23 0.48

20 3.03 0.50

21 1.89 0.45

22 2.71 0.48

23 3.31 0.47

Overall

While performance is not blistering, I am certainly not pining for my old provider. At EUR 30 per month with no line rental and a land line (over the Wi-Max network) thrown in it represents good value. I’ll probably upgrade to the 7MB product soon (they apply a charge for downgrades, but not to upgrades, hence my choice of a 3MB tariff at the outset).

There have been various moans and mutterings about the performance of this product, but as long as you are in an area of good coverage, you should be fine.

Follow up 2010-06-15: Three weeks on

After 21 days of pings, uploads and downloads I now have a reasonable amount of data to give a more reasonable picture. These graphs derive from data averaged over the 21 days. For the ping tests I averaged out the min, avg, max and stddev figures.

I stress (again!) that the server was not guaranteed exclusive access to the bandwidth and I know for sure that I was downloading hefty ISO files on more than one evening and there are other Wi-Fi users of the bandwidth too. This will affect all the results below.

While the graphs indicate very satisfactory performance, it is worth looking at the following table which illustrates the variability of the connection. In this case I have taken the overall averages for upload, download and pings. Without granting the test scripts exclusive access to the bandwidth it would be unfair to attribute any significance to the maximum ping time of over 2 seconds or the minimum download speed of 0.3 MBps – I can confirm that that figure refers to my hefty ISO download.

Overall averaged performance over 21 days

Down Up Ping

Min 0.30 0.14 29

Avg 3.42 0.41 87

Max 5.67 0.49 2402

StdDev 1.14 0.08 60

Graph of ping results (by time of day)

Graph of download/upload results (by time of day)

Hour	Min	Ave	Max	StdDev
0	76	166	1064	164
1	43	82	192	34
2	63	63	63	0
3	62	135	793	136
4	61	115	937	123
5	80	369	1644	342
6	61	93	221	36
7	46	90	262	36
8	32	65	103	10
9	76	83	128	16
10	74	149	1084	188
11	63	109	818	107
12	59	82	141	14
13	65	78	129	8
14	63	129	1307	176
15	60	103	525	79
16	46	82	158	16
17	31	83	194	33
18	54	81	143	19
19	51	86	197	28
20	40	168	759	140
21	76	258	1536	347
22	61	106	834	110
23	60	80	143	15

Hour	D/L	U/L
0	2.23	0.49
1	1.73	0.39
2	3.10	0.48
3	1.80	0.46
4	2.99	0.43
5	0.99	0.40
6	2.09	0.43
7	4.08	0.44
8	3.44	0.48
9	2.51	0.41
10	3.52	0.46
11	3.99	0.48
12	4.78	0.49
13	4.15	0.49
14	3.62	0.47
15	2.59	0.44
16	3.76	0.48
17	3.29	0.47
18	3.75	0.46
19	2.23	0.48
20	3.03	0.50
21	1.89	0.45
22	2.71	0.48
23	3.31	0.47

	Down	Up	Ping
Min	0.30	0.14	29
Avg	3.42	0.41	87
Max	5.67	0.49	2402
StdDev	1.14	0.08	60

Imagine Wi-Max Initial Review

admin — Tue, 11 May 2010 10:23:22 +0000

Last weekend I received a Wi-Max box from Imagine to replace my current DSL connection. Here are my initial thoughts on the device, the set-up and configuration, performance and some serious reservations about how Imagine have crippled the modem.

Delivery and packaging

I opted for the self-install option which is delivered by courier in a large jiffy envelope. Inside was an Imagine banded Motorola Wi-Max unit, a telephone handset adapter and a network cable.

The box itself is about the size of a hard-back novel (think a John Grisham airport edition), with a small swivel foot on the base to facilitate standing the device upright. It feels solid and well built. All antennae are internal, and there a series of four green LEDs on the front to show Wi-Max signal strength and one for Wi-Fi operation.

Connectivity: 1 x RJ-54 (Ethernet), 2 x Telecom (RJ-11), Wi-Fi. Disappointingly there is no hub built into the device.

Initial Set-up

Plug in and go (almost). According to the manual, as long as you have two of four signal strength LEDs lit, you are good to go. I had four, so I cannot comment on what performance would be like in marginally serviced areas.

Once up and running you plug in your PC (using the supplied network cable), open a web browser and you are automatically redirected to a set-up screen. From there you go through a series of five screens to confirm your PIN number (supplied by Imagine) and confirm your email address and activate the service. After that you are set up for Internet access. There is an additional step to activate and secure the Wi-Fi connection.

Annoyingly, the device kept dropping its’ Internet connection during the browser set-up phase and I was forced to start again from step 1. It took me several attempts to get through this step. This is a major irritation – I am in an area of high signal strength, so if you live in a marginally serviced area this step might get very frustrating.

Wi-Fi set-up was painless – just pick an SSID, a Wi-Fi radio channel and add a security password/key. However the limited access that Imagine give you to the device doesn’t really allow you to make a mistake here (See LAN Set-up).

Once I was through the set-up the device settled down and I was getting 7.5MB down and 0.5MB up with ping times of about 100ms. I’ll do a more rigorous test of this later.

I did not have a telephone handset to try out the VoIP service.

LAN Set-up

This is a major disappointment – if you want to customise anything for your local network, well, you just can’t. This includes setting your LAN network subnet (set to 192.168.15.x with no indication of what subnet mask is in use), port-forwarding, DMZ set-up, DHCP server settings.

I queried this with Imagine technical support and received two conflicting responses:

Access to the control panel and other features is restricted as changing some of these settings can cause the Wimax unit to cease functioning correctly.

and

There is proprietary technology there that we can’t let you see…

In my opinion both of these statements smell of bullshit – look at this review of Wi-Max and you can see screen shots of the Motorola browser interface.

What concerns me is that I see screen shots in that review for things like uPnP, firewall, enabling administrator access to the device (“Enable web login from Internet”) and port forwarding, but no indication from Imagine as to what the default settings are.

The rather idiotic response from Imagine is:

If you wish any ports forwarded please contact us either via email or phone and we can arrange that.

Seeing as you have very basic access to your own LAN using Imagine Wi-Max, this raises questions of liability if your network is compromised – especially in light of recent legal news over the “three strikes” agreement between Eircom and IMRA.

From a privacy and security standpoint there is no way of knowing what access to your LAN Imagine have given themselves.

Initial Conclusion

For a home user Imagine Wi-Max shows great promise. However the retarded attitude Imagine have towards the control panel, and the uncertain security set-up, make this of little use to the business or the technical user.