IIM Tutorial


Bandwidth and network speed exercise

  1. speedtest.net


The colleges network scales upon user amounts, the upload speed is expected to be relatively average but the download speed is, due to it being peak time, expected to be minimized for stress-handling– maybe even to a speed lower than the upload speed.

Over 802.11n, a current connection has a 90.19Mbps DL and 103.22 Up speed, resulting in a speedtest.net ping of 5ms. If I was connected over a wire, my DL speed would likely be faster but still limited by the bottleneck and scaling resulting from other hosts’ requirements.

  1. Using the PCs clock or your watch, evaluate the downloading of a large image

The image file is 1.15 MB in size, roughly 9Mb, the image downloaded in a second, appropriate for the relatively fast 90Mbps (11MBps) DL speed connection in use. In actuality, the file downloaded in around a tenth of a second – a timeframe difference unable to be seen with the naked eye.


The second’s worth of transmissions included networking on different layers, with some time devoted to physical linking, data-link transmission, application execution and session authentication.

  1. Did the response time match the speedtest.net bandwidth figure? If not, what might the reason(s) be?

The response time was in accordance with the polled transfer rates, as in the transmission time did not take longer than the rate dictates. In solid terms of human perception, there was no response time mismatch.

  1. Using the suitable program, save the image in different formats – e.g. jpg, bmp, jpg, gif and png. Compare the compression ratios.

Outside of traditional dedicated compression software, one method to change file sizes of images is the use of Paint and saving in another file format. Additionally, one can resize the image to a smaller area of pixels and have increased control. Paint, by default, offers to save images in Jpg, Bmp, Bmp or GIF (typically used for image sequencing).


The image resolution of 1300 x 1000px did not change, yet the file size and image quality did. With precisely the same properties, only the bit formats are changed. Applications, Word in th, treat the same 60x60px sized slices differently depending on file format, displaying them at different sizes by default, with smaller file sizes generally displayed larger reliant on compression technique.



  1. About video and audio with different file types.


MPEG Video formatting is dependent on multiple metrics.

Varying-sized Bitrates are typically comprised of a joining of resolution, frame rate, interlaced/progressive rendering type, audio format+bandwidth, compression method and color palette componential variants.

Development of H/X265 over traditional H/X264, results in smaller file sizes at around 2x higher processing requirements.

SOURCE: 8.30MB MP4 (640×360)

21.2 MB
WMV 20.8 MB
AVI 18.6 MB
MKV 10.1 MB
MP4 10.1 MB
FLV 10.1 MB
MOV 9.43 MB


FLAC Audio formatting is primarily dependent on bitrate and encoding method.

3GP file is so small as the bitrate is 31kbps

FLAC is the largest as bitrate multiplied roughly 4x to 761 kbps

SOURCE: MP4 (length 1:06:06) 194kbps

360 MB
MP3 91.8 MB
AAC 75.6 MB
3GP 15.2 MB


  1. Network used by students of FW-208

On assumptive visualization: Access switch area (classroom) to MLS on upper floor to router in other room down to grounded ISP.


  1. Guestimates of the traffic

With around 40 students visible (henceforth requiring bandwidth) and sufficient download speeds being around 250kBps, serving responsive interaction for current general purpose downloading, 1000kB is estimated required for satisfying service to 4 hosts and 10,000kB (8Mb) bandwidth for 40.

The single-host speedtest.net DL result of 90Mbps shows that the college infrastructure’s existing capabilities are currently more than sufficient and somewhat ‘future proofed’, expectedly using a PRI or fiber wire public connection.

  1. If the university has 1GB connection.

8Gb (1GB) bandwidth is 1024MB (8,192Mb), 91 times the single-host DL-speedtest and 1024 times the 40-host classroom congestion guestimate. Therefore, one can estimate that the infrastructure’s public connection is capable of handling around 40,000 student connections worth of traffic.

This high bandwidth-capability is entirely dependent on the capabilities of supporting devices in the infrastructure, e.g. the AAA service and the distribution switching, in entirety, would need to be able to handle 40,000 users – allocating supernet addresses and performing address translation without overrun discards. In consideration of expected demand, the 1GB connection is sufficient for general use but, under maximal demand, is prone to bottlenecking and scaling access speeds down to regular demand user-crippling rates.



Leave a Comment...

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s