1. General

2B2T is the Monitoring and Evaluation framework for broadband connections. Only registered users may view a history of their connection statistics. Registered users contribute to the aggregated broadband statistics for Cyprus.

2B2T users may test the quality of their broadband connection, via the Network Diagnostic Tool. Connection statistics per user and aggregate statistics per postal area/municipality/county/region are displayed with maps, graphs and lists. Moreover, registered users have the ability to manage their 2B2T account. A detailed description of the 2B2T functionality is presented in the following paragraphs.

2. Introduction to 2B2T

Each 2B2T web page is divided into three sections. Section 1 is presented in the following figure.

To change the language (Greek/English) click on the language tab on the top left side of the section. On the top right side of the section, the authenticated user's e-mail is displayed along with the Settings the account. Help and Login/Logout links are also available. 2B2T logo is an active link to the 2B2T homepage.

Section 2 displays the measurement tools frame (NDT can be subsequently selected), maps and graphs. Section 2 is presented in the following figure.

The tabs at the top of section 2, allow the selection of 1.the measurement tools' page, which contains NDT, 2. the map of Cyprus with various regions colored according to the corresponding user measurements and 3. the graphs of broadband measurement results aggregated through time and space for the authenticated user.

The last section of the page provides useful links and information regarding SPEBs. Section 3 is presented in the following figure.

3. User accounts

Every registered user may login to 2B2T according to the following procedure:

Registered users must provide their email address along with the password set at the time of registration. Optionally the "Remember me" checkbox may be used to avoid filling in the username/password in subsequent visits. A user may request a new password via the Forgot my password link. The user must provide the registered email address in order to receive a new randomly generated password. Users can change their password in the settings page.

Users may select Create new account or Settings in order to provide for the first time or alter personal information.

The first set of attributes correspond to personal user information. Users must provide the password twice, to ensure that it has been spelled correctly. Appropriate system messages guide users through the registration process. The user can also opt in to receive news regarding the service from the Hellenic Telecommunications Commission.

The rest of the fields are attributed to the user's broadband connection. .

The next section of the form must be filled with geographical information regarding a user's broadband connection. Specifically users must provide a valid postal address corresponding to their broadband connection, in order to render statistics per postal area/municipality/county/region. Auto-complete capability is provided regarding the postal code and municipality information. Based on provided information, a user's postal address is displayed on the map.

At the last section, users may define the settings of the purchased broadband service (ISP, nominal access data rate).

You must agree to the terms and conditions and type the two words that appear in the text box provided, to prove that you are an actual person.

Select Create Account to complete the registration process. An email is sent to the email address set by the user during registration. The email contains a link that will activate the account. By selecting Return the registration process is cancelled.

Registered users can use the form to alter their personal information. However, in this case password update is not required. If the user chooses to change information related to the broadband connection or the postal address, a new broadband connection is registered at the user's account. Every user is associated to only one active broadband connection at a time.

4. Perform Measurement - Network Diagnostic Tool

Network Diagnostic Tool's primary purpose is to carry out end to end measurements on residential broadband connections as it incorporates functionality to approximately identify characteristics and evaluate performance of the user's broadband connection. Some of the most interesting network quality metrics observed by the Network Diagnostic Tool are upstream/downstream bandwidth (upload/download speed) to the user's pc, Round Trip Time (ping time) as well as the variation in the RTT (jitter) of individual packets and the percentage of packet loss. The architecture of NDT is based on the interaction of two distinct peers (a web browser acting as the measurement client and a measurement server) that evaluate the quality characteristics of the network path between them. The basic idea of this interaction is the following; the client and the measurement server exchange a predefined amount of network traffic and, based on timings of the exchanged packets the server calculates certain metrics pertaining to the quality of the link between them.

A modified version of the NDT HTML5 Client has been adopted, in order to simplify the presentation of the performance evaluation results for the user. The functionality of the measuring instrumentation has been left unchanged. Users may perform advanced NDT measurements. The modified NDT renders the measurement results at the web client in appropriate format allowing for enhanced presentation of broadband connection statistics.

4.1 NDT usage

Following the appearance of the NDT measurement page, the user may select Start Measurement. The measurement process is comprised of several distinct stages that can be identified at the progress bar. The basic measurement metrics such as upstream/downstream bandwidth (upload/download speed), RTT (ping time), RTT variation (jitter) and packet loss, are registered during the completion of the various measurement stages. Average execution time of a typical NDT measurement varies from 30 to 60 seconds. The indication Measurement is complete at the progress bar follows the completion of every measurement. The end user at this stage can repeat the measurement process.

4.2 Performance evaluation metrics
  1. Upstream/Downstream Bandwidth (Upload/Download speed) The specific metric provides an indication of the available bandwidth at the end user's pc. Bandwidth estimation is based on the entire client/server communication path. Increased downstream bandwidth allows for increased rate downloading. Along a communication path, the maximum traffic rate between the client and the server is constrained by the most congested link. Usually the most congested link for a typical ADSL end user is the access link to the broadband service provider. However, congestion on backhaul links along the communication path is also possible and will be reflected at the measurement.
  2. Round Trip Time (Ping time) The RTT is the time needed for a packet originating from the broadband host to the measurement server increased by the time needed for the server to respond to the host. Regarding RTT calculation both the broadband host and the measurement server can be used as origin/destination irrespectively. RTT is a very significant performance evaluation metric. An increased value of RTT could lead to increased delay for applications such as interactive network games, Internet telephony (VoIP, Skype etc). The average value off RTT is calculated from a predefined number of tests on both directions.
  3. Packet Loss Packet loss is calculated along the round trip paths from the MLAB measurement server to remote broadband hosts. Values around 1% indicate serious disruptions in the service.
  4. RTT variation (Jitter) RTT variation measures how spread are the RTT values around the average e.g. an RTT variation of 3 milliseconds indicates the difference between minimum and maximum calculated RTT values. The metric is also an indicator of possible disruptions in the service, however it must be noted that since extreme values are taken into consideration it is not a safe performance evaluation metric.
5. Measurement results and aggregate statistics

Measurement results are available in three versions. The user may view and save measurement results as a detailed list and graphs in time. Moreover the user has the option to view aggregate measurement statistics on the map of Greece.

5.1 Map

Measurement statistics map is displaying the results of the statistical analysis per region. Authenticated users may be notified of measurements in their surrounding area.Non authenticated users may view aggregate statistics for Greece. Depending on the zoom level, measurements are aggregated over space in different resolutions. From lower to higher resolution, measurements are aggregated over region, county, municipality, postal area.

Each polygon is an abstract representation of an area. Polygons are symmetrical around the centre of the area. A polygon is coloured according to the average value of the metric associated with it. To select a specific metric, click on the corresponding tab. The quality characteristics that can be seen on the map per measurement tool are:

For the Network Diagnostic Tool, the following quality characteristics can be displayed:

  • Downstream bandwidth
  • Upstream bandwidth
  • Round Trip Time
  • Packet loss
  • Jitter
The values for each of the abovementioned indicators are divided into suitably selected ranges, as indicated in the colored legend right bellow the map. The color of each polygon is decided according to the range which includes the calculated average value for the selected characteristic. The selection of the characteristic can be done by using the appropriate labels right above the map. Clicking a polygon activates a balloon with all the average values for the selected tool, regardless of the selected displayed characteristic.

Every broadband connection in a area contributes to the average values. You can exclude a specific ISP from the aggregate statistics by deselecting the ISP from the list on the right side of the window. Polygon colouring is automatically adjusted.

Google Maps API has been used to create the map. Therefore, zoom and move events are supported. Users may utilize the mouse or the zoom control slider for that purpose. Moreover different map layers are supported (map, satellite, terrain).

In a higher resolution, measurement statistics are available per broadband connection. In addition, the following indicators are present along with the selected tool measurements for each broadband connection:

  • Distance from the nearest exchange
  • Theoretical maximum attainable ADSL speed based on the abovementioned distance
  • Theoretical maximum attainable VDSL speed based on the abovementioned distance
  • Headline speed for the broadband connection
  • Number of measurments carried out

Average values are calculated per broadband connection, using registered measurement results. The broadband connection of the authenticated user is depicted with a larger icon. The same metrics and colouring scheme applies as in the case of polygons. You can exclude a specific ISP from the individual statistics by deselecting the ISP from the list on the right side of the window. By clicking on a specific icon, average values for the five metrics appear regarding this broadband connection.

5.2 Graphs

Authenticated users may register their measurement results. These values are used to calculate broadband connection statistics. The user has also the possibility, via the available graphs, to view the history of measurements performed on his/her broadband connection.

Measurement results are presented as a function of time, on a daily basis. In the case that multiple measurements have occurred per day, then the average value is calculated. At least three measurements are required for graphic display of results.

Downstream bandwidth (download speed), upstream bandwidth (upload speed), RTT (ping time), packet loss and RTT variation (jitter) measurements are presented. Downstream and upstream bandwidth are displayed on the same graph. Dragging the pointer over a graph line, displays the date and value of the selected metric at the top right corner of the graph.

5.3 Measurements list

Measurement results are presented after the completion of the selected tool measurement process. The user's measurements history contains all registered measurements regarding the specific connection. The history is available by clicking on the tab Details.

Each measurement performed by a registered user has a specific timestamp (date and time). By clicking on Export data as a CVS file users have the ability to save the measurements results on their local PC. The file is editable e.g. via Microsoft Office Excel.

6. Frequently asked questions (FAQ)

Ε1: How accurate are the NDT measurement results?
Α1: The NDT tool provides an approximation of the measured metrics. The quality of the estimated values can be affected by other applications running at the user's PC or at the LAN as well as temporary problems encountered on the network path to the server. The user is advised to perform the measurement on a idle broadband connection.

Broadband measurement tools selected by EETT provide an estimation of broadband connections quality. The results are presented impartially with the goal to inform broadband subscribers on the market of broadband connections. EETT can not guaranty the accuracy of the measurements that depends on multiple external factors beyond the control of the national regulation authority. The user is responsible for correct and sufficient use of the provided tools. EETT is not responsible for any problems that may arise from the use of the broadband measurement tools.

Ε2: Can the measurement be performed from a WiFi connected PC?
Α2: Depending on the 802.11 flavour, the bandwidth is constrained by the bandwidth of the wireless medium. In the case of 802.11b, depending on the signal strength, bandwidth varies from 11Mbps, 5.5Mbps, 2Mbps or 1Mbps. Moreover, due to the wireless environment the measurement results may indicate that a half duplex connection is used. Therefore it is suggested to perform the measurement on a PC directly connected to the modem/router.

Ε3: Which are the parameters that affect the quality of a measurement?
Α3: Firewalls filtering traffic, high load at the measurement server or cabling related problems may affect the quality of the broadband connection.

Ε4: Do the measurement results include data overhead (eg from TCP/IP headers)?
Α4: No, bandwidth value is not calculated with protocol overhead.

Ε5: Can the nominal data rate be achieved? For example if the user purchases a 24 Mbps DSL service, can the results of the measurement (download speed) reach that value?
A5: Every utilized protocol (TCP,IP,PPP etc) adds overhead to the data transfer. Therefore measurement results are always less than the nominal data rate.

Ε6: What is the overhead in data transfer over a broadband line?
A6: It depends on the specifics of the broadband connection (e.g. protocols used).

Ε7: Which is the transport protocol used by NDT?
A7: NDT utilizes TCP as implemented by the server. TCP connections are established between the client and the measurements server.

Ε8: How is RTT, packet loss and RTT variation calculated?
Α8: NDT utilizes TCP RTT estimation function (Van Jacobson: Congestion avoidance and control. SIGCOMM 1988).Jitter calculation is based on the TCP RTT estimation and is defined as the difference between extreme RTT values. Retransmited packets contribute to packet loss estimation.

E9: How is the nearest local exchange determined?
A9: For urban areas in Athens and Salonika, the exact polygons surrounding the area serviced by each exchange are used. Consequently, every connection can be safely assigned to the exchange that is actually being used. For the rest of the country, selection of the nearest exchange is hitherto based on the absolute geographocal distance.

E10:How is the distance to the nearest exchange calculated?
A10:The approximate distance to the nearest exchange is calculated according to the geographical coordinates of the exchange and the user connection location. Cable lines usually track the existing road infrastructure, thus the distance is not calculated based on the geographical distance but is approximated by summing the vertical and horizontal differences between the two locations.

E11: How is the the maximum theoretical speed calculated for ADSL and VDSL connections?
A11: The maximum theoretical distance is calculated according to the following tables:

< 300m23.5 Mbps
< 600m22.5 Mbps
< 900m22 Mbps
< 1200m20 Mbps
< 1500m18 Mbps
< 1800m17 Mbps
< 2100m15 Mbps
< 2400m14 Mbps
< 2700m12 Mbps
< 3000m9.5 Mbps
< 3300m8 Mbps
< 3600m6.5 Mbps
< 3900m5 Mbps
< 4200m4 Mbps
< 4500m3.5 Mbps
< 4800m2.5 Mbps
< 5100m2 Mbps
< 5400m1.5 Mbps
< 5700m1 Mbps
< 6000m768 Kbps
< 6300m512 Kbps
< 6600m128 Kbps

DistanceVDSL upstream VDSL downstream
< 250m8 Mbps 42 Mbps
< 500m7 Mbps 36 Mbps
< 750m6 Mbps 33 Mbps
< 1000m5 Mbps 28 Mbps
< 1250m4 Mbps 21 Mbps
< 1500m1 Mbps 17 Mbps