Expert recommends setting up network standard server

"Chengmenthe implicating. "The broadband speed is false, the speed test software naturally also take off the relationship."

Since the CCTV burst "false broadband truth", as the phenomenon of "accomplice" test software has been pushed on the cusp, about the speed of the software why the interpretation of uncertainty is also frequently on camera. In fact, not only is the test results to be skeptical, different test software to measure the diversity of data also makes people confused.

Different speeds of the same network

Recently, the reporter used different speed-measuring software, in 10 minutes to their own computer to speed. The test results are all the same in the case where the program used and the website visited remain unchanged.

The reporter first used 360 broadband gun, test results show: The reporter used the computer download speed reached 815kb/seconds, ahead of the national 93% users; and according to another interesting network in the speed test software testing results: The computer download rate of 456.62kb/seconds, Upload speed of 680.27kb/seconds, more than 63% of the network speed of 4 stars, the test results show bandwidth of 2M; however, the test results of another eight-week studio software are quite different from the two software: computer download speed is 206.7kb/seconds, Upload speed 24.63kb/seconds, network speed level of Samsung, network bandwidth of 3M.

Several other network test software results are: ADA software, upload speed 244kb/seconds, connectivity rate of 81%, speed beat the national 77% test users; Web application software, download speed of about 961.54kb/seconds, upload speed of about 227.79kb/seconds.

Each software test results are different, and not only the bandwidth is different, which upload and download speed also has a huge difference, the largest upload speed 680.27kb/seconds and the smallest upload speed 24.63kb/seconds, with close to 30 times times the gap.

Why inconsistent

Regardless of the speed of the different test software test results between the "fight" situation how to explain it? Fu Liang, vice president of the network accelerator provider Blue Flood, explains one reason: "There are a lot of flash speed software on the Internet, most of them are single channel, the computer and the server to create a single link between the download a file, and then tell you how much speed, this way is the most unscientific. ”

However, there have been the speed of the software to respond to similar queries, such as 360 in response to CCTV reports, said the 360 broadband gun working principle is not adopted analyzes method, but the use of HTTP multi-threaded download and wide-area network Peer-to-peer download method, is the most scientific method of speed measurement.

In this respect, Beijing 3G Industry Alliance vice President Xiangli just gave another explanation, due to different speed measurement software between the speed mechanism, resulting in a number of test results between the differences.

"If the test software and network provider in the same area, using the same operator, the same server, the test is equivalent to the same LAN, the results will be more ideal, and the contrary, the operator and the server is different, the results will be lower." "The case has just used an image example," If a network using Beijing Unicom's server, and testing software used by the server is Shenzhen Telecom, software testing will be around a large circle, the results will certainly be low. ”

In addition, the speed of the software selected "Network segment" is also about the test results of one of the reasons. According to Longwen, director of the Interactive Marketing Research Institute introduce introduced, from the user home routers to the operator network throughout the backbone network, will affect the user's network experience, but the speed of the software is often only selected in one of the network segment for testing, so that the test results inconsistent.

How can the speed measurement be accurate?

The speed of the software so unreliable, then consumers spend the net to buy the speed, how can there be a clear statement? According to Xiang, currently there is no national standard for testing software, so the consistency and accuracy of test results is unlikely to be achieved.

It is understood that the Ministry of Industry in April this year has publicized the "fixed broadband access rate test Method" industrial standards for the approval of the draft, defined the "fixed broadband access rate" concept, and the test equipment put forward requirements, set the benchmark test methods, alternative testing methods and test data processing methods. At present, the industry standard has not yet been implemented, but has been high hopes for many parties.

"In addition to a unified testing approach, a standard server is required." "Xiang that the use of different testing software to go through different servers, so the test results will vary, want to obtain a unified network test environment, you need to have a national standard server, as a unified access point for each network." "China Quality News"