Free Essay

Benchmarking of Wireless Routers

In: Other Topics

Submitted By pjbustamante
Words 4964
Pages 20
Network Performance Project
Benchmarking of Wireless Routers
Pedro Bustamante

Noopur Chowdhary

School of Information Science
University of Pittsburgh
Pittsburgh, United States pjb63@pitt.edu School of Information Science
University of Pittsburgh
Pittsburgh, United States noc13@pitt.edu Abstract— This paper explores a basic technique for Network
Performance Analysis: Compare alternate equipment based on real measurements. Further, this work describes the comparison between two wireless routers configured as access points. The analysis of the equipment was implemented in terms of throughput (TCP) and jitter (UDP). In this way, this work aims to show the results of the benchmarking process of two wireless routers from different vendors through the utilization of a basic performance terminal tool such as “iperf”.
Keywords—benchmarking;
wireless; routers.

network;

performance;

II. GOALS


First, this project seeks to execute, analyze and evaluate a Network Performance Analysis based on real measurements. •

To benchmark two wireless routers of different vendors using an active software network performance tool such as iperf.



To measure and evaluate the total available throughput of parallel TCP connections over tested wireless routers
(acces points) in a server/client configuration.



To determine and exanimate the average jitter and percentage of lost packets of parallel UDP connections over the tested wireless routers (acces points) in a server/client architecture.

iperf;

I. INTRODUCTION
There is no doubt that Wi-Fi (IEEE 802.11), the most globalized Wireless Network LAN, has become one of the most noticeable ways to connect devices in universities, homes, business, etc. Nevertheless, not many users know about the complex performance factors of a Wi-Fi network, and how all the devices are able to stay connected exchanging information with the world. Moreover, the performance factors of the key element of the network: the wireless router (access point).
In this way, in the technical field (leaving the cost aside, which is most of the time the considered factor) signal strength is often the in effect performance benchmark quality of a wireless router (i.e., performance of the network).
Nonetheless, for a deeper analysis signal strength alone could not be enough to determine the viability and usability of a network. All too often, a network’s signal strength will fall within acceptable ranges, yet have insufficient data rates for an optimal (at least usable) connectivity. Considering this fact, this project aims to analyze other characteristics of wireless routers such as Throughput (TCP) and Jitter (UDP).
Indeed, this experiment setup and analysis presents the results of the comparison of two wireless routers of different vendors. The analysis of the devices would be centered in key characteristics of the most common transport layer protocols used nowadays in the internet: Transport Control Protocol, and
User Datagram Protocol.

III. GENERAL INFORMATION
A. Wireless Routers
As observed the present work is based in the benchmarking process of two wireless routers. Thus, Table I details the characteristics of the equipment to be compared.
TABLE I.
Routers
Router 1
(R1)
Router 2
(R2)

WIRELESS ROUTERS

Characteristics
Brand

Model

Netgear

WNDR4300 802.11N Dual Band

Linksys

WRT600N 802.11N Dual Band

The compared routers are commercial residential devices. In addition, they posses similar technical features (processor speed and RAM) to guarantee an objective evaluation. Further, the chosen routers are designed for medium to large residential spaces and to operate under the 802.11n standard. Besides, both devices are capable of reaching speeds between 300 to 450
Mbps (Information gathered from the vendor webpage).

Besides, both routers use the latest available property firmware, and were configured as access points to serve as a link in a client/server architecture. Consequently, the chosen wireless routers are configured undre normal parameters (no added functionalities or firmware modifications) to obtain the best possible performance under normal enviromental conditions.

Factors
TCP
Connections
UDP
Connections

Description

2

2 parallel streams / 5 parallel streams

2

2 parallel streams / 5 parallel streams

IV. METHODOLOGY

B. Used Tools
Due to the nature of this project, a tool with not only traffic generation but also measurement capability was needed.
Therefore, the natural choice was: Iperf. This performance tool is a widely-utilized open-source active software performance tool written in C. Besides, Iperf 3 (latest version) runs on several operating systems using the command window (or terminal) to test and/or evaluate a Network. It allows the creation of TCP (Transmission Control Protocol) and UDP
(User Datagram Protocol) data streams between two hosts in a
Client/Server architecture. The basic functionality of the tool is to measure the throughput (Bits/unit of time), jitter (rate of change in the end-to-end delay) and number of lost packets of the tested network (percentage).
It is necessary to point out that all the set of results were saved in text files (.txt) to keep record of the complete measurements and to facilitate the post-processing tasks.
Indeed, all the original files were processed and analyzed using
MATLAB ® and Microsoft Excel ®.
C. Factors and Variables
The first step in the experiment design is to choose the adequate response variables. In this way, this project will utilize two main response variables to analyze the performance of each of the wireless routers:



Factors and Levels for the Experiment
Levels

A. Experiment Design
As shown in Table II, 5 experimental factors with 2 levels each were chosen for the benchmarking task of this project. For each combination of factor and level, the experiment recollected 100 values of the response variables (replications).
Furthermore, the Total Number of Experiments (TNR) was obtained using the Full Factorial Design (all combinations of levels and factors); which resulted in a total of 3200 recollected values in the experiment divided in 32 sets of measurements. = ∗

)
(*+

(

= 10 22222 = 3200

(1)
(2)

B. Data collection
As explored in the previous sections, the testing was aimed under a Full Factorial Design. In this way, all the factors and its levels were combined. For instance, to obtain a determine throughput we could have: Router: 1, Frequency: 2.4, TCP, 2 parallel connections and a distance of 150ft in a set of measurements. For each of the tests (32 in total), the routers were connected as access points, serving as the link between the iperf Client (traffic generator) and the iperf Server (point of measurement) as it is shown in Fig. 1.

Throughput: Measured using parallel TCP connections
End-to-end delay variation (Jitter): Measured using parallel UDP connections.

Additionally, to guarantee a zero impact of external factors in the measurements the signal characteristics of the routers were monitored during the complete experiment.
On the other hand, the factors (parameters to be varied in the study) and its levels (values for the factors) of the experiment were also selected (Table II).
TABLE II.
Factors

FACTORS AND LEVELS

Factors and Levels for the Experiment
Levels

Description

Routers

2

Netgear / Linksys

Frequencies

2

2.4 GHz / 5.0 GHz

Distance

2

20 Feet / 150 Feet

Fig. 1. Experiment Architecture

The iperf client was configured in a single machine (Apple personal computer). In the same light, the iperf server was configured in another independent PC (Apple). Both computers had similar and sufficient technical characteristics (Operating system, processor speed, hard drive and bus capacity), which minimize any impact of the end-hosts in the results. In addition, each replication measured the throughput for each parallel stream and a resulting throughput (sum of the individual

throughputs). To fulfill the designed experiment setup, iperf
(version 3) was configured as follows:

in terms of Signal Strength. Consequently, no future analysis will be done respecting this matter.

iperf -c 192.168.1.1 -P 1 -i 1 -p 5001 -C -f m -t 100 -T 1
(Client)
iperf -s -P 0 -i 1 -p 5001 -C -f m
(Server)

The UDP configuration of the project was done in the same manner as TCP: generating traffic from client to server using the aforementioned equipment. In addition, each replication measured jitter and percentage of lost packets for each parallel stream and a resulting throughput (sum of the individual throughputs). Thus, we have the UDP configuration as follows: iperf -c 192.168.1.1 -u -P 1 -i 1 -p 5001 -C -f m -b 1.0M -t 100 -T 1
(Client)
iperf -s -u -P 0 -i 1 -p 5001 -C -f m
(Server)

C. Sets of Measurements
Table III details the combinations that were tested using the above configuration of the tool (Same for distance = 100 feet and Router 2 as well).
TABLE III.

Fig. 2. Signal Analysis (SNR (Quality) and Signal Strength)

E. Benchmarking
To analyze the individual performance of each router and its posterior comparison against each other, the following procedure was implemented:
1.

COMBINATIONS OF FACTORS (SETS OF MEASUREMENTS)

Combinations of Factors

Combination

2.

Description

1

Router 1/2.4 GHz/TCP/2 Parallel Connections/20ft

2

Router 1/2.4 GHz/TCP/5 Parallel Connections/20ft

3

Router 1/5.0 GHz/TCP/2 Parallel Connections/20ft

4

Router 1/5.0 GHz/TCP/5 Parallel Connections/20ft

5

Router 1/2.4 GHz/UDP/2 Parallel Connections/20ft

6

Router 1/2.4 GHz/UDP/5 Parallel Connections/20ft

7

Router 1/5.0 GHz/UDP/2 Parallel Connections/20ft

8

Router 1/5.0 GHz/UDP/5 Parallel Connections/20ft

3.

= = ± ∗

D. Signal Characteristics
Finally, it is necessary to point out that in each of the tests
(32 in total), the signal strength of the routers was monitored all the time. For this purpose, the embedded Apple tool “Wireless
Diagnostics Performance” was utilized. The signal characteristics that were monitored were the Signal Strength in
(dBm) as well as the Signal to Noise Ratio (dB), as shown in the examples in Fig. 2.
From this monitoring process, we can conclude that for none of the 32 tests the SNR of the Signal Strength were as critical as to be considered a constrain in the measurements. In other words, both routers displayed an acceptable performance

Verify if the measured data was correlated. Further, the coefficient of correlation was calculated for each set of measurements (32 in total)
The present project used measurements of central tendency for comparing purposes. Indeed, the following values were calculated for the analysis:
a. Total average throughput under the TCP configuration (Sum of the individual streams) b. Average jitter using the UDP configuration
(Average of the individual streams)
c. Average percentage of lost packets for the
UDP tests
Using the aforementioned values, a 90% confidence interval was calculated for each of set of measurements (32). Each of the CI was obtained through the following expression:
@
A

= 1.645 = ; = . ; = #

4.

(4)
(5)
(6)

Once a CI was calculated for each of the sets of measurements, a comparison (benchmarking) between the routers was possible. Thus, when there is no overlapping in the CI, a conclusion can be made in terms of a better performance metric of one wireless router over the other

V. RESULTS

3
2.5

A. Original Information TCP

2
1.5
1
0.5
0

1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97

As mentioned in the previous section, the TCP tests were implemented using two variations: 2 and 5 parallel streams.
Moreover, for each test the effective throughput was presented as the individual measurements and the sum of the signals (Fig.
3 and Fig. 4).

Stream 1
16

Stream 2

Fig. 5. Jitter Measurements Example (R1/2.4/UDP/2/150ft)

14
12

35

10
8

30

6

25

4

20

2

15

SUM

Stream 1

Stream 2

Fig. 3. Throughput Measurements Example (R1/2.4/TCP/2/150ft)

10
5
0

1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97

1
4
7
10
13
16
19
22
25
28
31
34
37
40
43
46
49
52
55
58
61
64
67
70
73
76
79
82
85
88
91
94
97
100

0

Stream 1

40

Stream 2

Stream 3

Stream 4

Stream 5

Fig. 6. Jitter Measurements Example (R1/2.4/UDP/5/150ft)

35

Additionally, under the UDP testing scheme, iperf also measures the percentage of lost packets (lost packets/total packets) of each stream in the connection. This characteristic is shown in Fig. 7.

30
25
20
15
10

120%

5

100%
1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97

0
SUM

Stream 1

Stream 2

Stream 3

Stream 4

Stream 5

80%
60%
40%

B. Original Information UDP
The UDP measurements were configured in the same manner as the TCP tests (2 and 5 parallel streams).
Nevertheless, the response variable to analyze was jitter through the individual jitter of each stream.

20%
0%

1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97
101

Fig. 4. Throughput Measurements Example (R1/2.4/TCP/5/150ft)

Stream 1

Stream 2

Stream 3

Stream 4

Stream 5

Fig. 7. Packet Lost Measurements Example (R1/2.4/UDP/5/150ft)

C. Results Router 1
All the measured information was processed for each of the routers as the first step in the comparison process. In this manner, Table IV shows the average results in terms of throughput obtained after each TCP test (8 in total) of the first benchmarked router (Netgear WNDR4300).

TABLE IV.
Frequency
(MHz)
2.4
2.4
5.0
5.0
2.4
2.4
5.0
5.0

Protocol
TCP
TCP
TCP
TCP
TCP
TCP
TCP
TCP

RESULTS ROUTER 1(THROUGHPUT)
Router 1 (Netgear)
Parallel
Distance (ft)
Connections
2
150
5
150
2
150
5
150
2
20
5
20
2
20
5
20

7.0000
6.0000

Average Throughput
(Mbits/sec)
9.6736
5.7620
39.6083
36.7984
42.8203
44.6169
100.1848
103.8958

5.0000
4.0000
3.0000
2.0000
1.0000
0.0000
1
2.4 UDP 2 150 2.4 UDP 5 150 5.0 UDP 2 150 5.0 UDP 5 150

120.00

2.4 UDP 2 20

2.4 UDP 5 20

5.0 UDP 2 20

5.0 UDP 5 20

Fig. 9. Jitter Results Router 1 (Table V)

100.00
80.00
50.00%

60.00

45.00%
40.00%

40.00

35.00%
30.00%

20.00

25.00%
20.00%

0.00

15.00%

1

10.00%
5.00%

2.4 TCP 2 150 2.4 TCP 5 150 5.0 TCP 2 150 5.0 TCP 5 150
2.4 TCP 2 20

2.4 TCP 5 20

5.0 TCP 2 20

0.00%
1

5.0 TCP 5 20

2.4 UDP 2 150 2.4 UDP 5 150 5.0 UDP 2 150 5.0 UDP 5 150

Fig. 8. Throughput Results Router 1 (Table IV)

2.4 UDP 2 20

In the same light, Table V and VI expose the average jitter and percentage of lost packets. This values were obtained for the different UDP performance tests (8 in total) of the Netgear wireless router (R1).
TABLE V.

RESULTS ROUTER 1(JITTER)

Parallel
Connections

Distance (ft)

Average
Jitter
(msec)

2.4
2.4
5.0
5.0
2.4
2.4
5.0
5.0

UDP
UDP
UDP
UDP
UDP
UDP
UDP
UDP

2
5
2
5
2
5
2
5

150
150
150
150
20
20
20
20

1.3685
6.2372
0.5872
1.8602
1.4564
4.2698
0.3693
0.5581

TABLE VI.
Frequency
(MHz)
2.4
2.4
2.4
2.4
5.0
5.0
5.0
5.0

Protocol
UDP
UDP
UDP
UDP
UDP
UDP
UDP
UDP

RESULTS ROUTER 1(PACKET LOSS)
Router 1 (Netgear)
Parallel
Distance (ft)
Connections
2
150
2
150
5
150
5
150
2
20
2
20
5
20
5
20

5.0 UDP 5 20

From the aforementioned information we can describe the performance of the first wireless router as follows:


Protocol

5.0 UDP 2 20

Fig. 10. Packet Loss Results Router 1 (Table VI)

Router 1 (Netgear)
Frequency
(MHz)

2.4 UDP 5 20



In terms of throughput we can observe that at less distance
(router to server/client) a higher throughput is obtained. In the same way the 5.0 GHz frequency of the equipment displays a higher throughput rate.
As expected, in the UDP measurements, the amount of jitter and number of lost packets is directly impacted by the number of parallel streams and the distance. Thus, more streams and distance are translated in a bigger difference in the end-to-end delay (jitter) and in more loss of packets between client and server.

D. Results Router 2
Packet Loss
(%)
0.16%
47.37%
2.26%
9.02%
0.00%
0.00%
30.43%
47.55%

Table VII shows the average results in terms of throughput obtained after each TCP test (8 in total) of the second benchmarked wireless router (Linksys WRT600N).
TABLE VII.
Frequency
(MHz)
2.4
2.4
5.0
5.0
2.4
2.4
5.0
5.0

Protocol
TCP
TCP
TCP
TCP
TCP
TCP
TCP
TCP

RESULTS ROUTER 2(THROUGHPUT)
Router 2 (Linksys)
Parallel
Connections
2
5
2
5
2
5
2
5

Distance (ft)
150
150
150
150
20
20
20
20

Throughput
(Mbps/sec)
5.3072
6.3169
5.5941
5.7859
8.9916
8.9916
6.9441
7.1983

0.00

2.00

4.00

5.0 TCP 5 20

6.00

5.0 TCP 2 20

8.00

2.4 TCP 5 20

0.00%

10.00

20.00%
5.0 UDP 5 20

2.4 TCP 2 20

Fig. 11. Throughput Results Router 2 (Table VII)

Frequency
(MHz)
2.4
2.4
5.0
5.0
2.4
2.4
5.0
5.0

Protocol
UDP
UDP
UDP
UDP
UDP
UDP
UDP
UDP

TABLE IX.
Frequency
(MHz)
2.4
2.4
5.0
5.0
2.4
2.4
5.0
5.0

Protocol
UDP
UDP
UDP
UDP
UDP
UDP
UDP
UDP

2.4 UDP 5 20

100.00%

2.4 UDP 2 20

From the aforementioned data we can describe the performance of the second wireless router in the following terms: RESULTS ROUTER 2(JITTER)

In the TCP configuration we have that: we can have more throughput rate at smaller distances. In the same way the
5.0 GHz frequency presents a higher throughput than the
2.4 band in the same router.
For the UDP measurements, the amount of jitter and number of lost packets is directly impacted by the number of parallel streams and the distance. Thus, more streams and distance are translated in a greater variation in the endto-end delay (jitter)


Distance (ft)

Jitter (msec)

150
150
150
150
20
20
20
20

4.5386
8.7997
5.8742
15.0149
1.7606
4.0174
4.2845
5.3822



RESULTS ROUTER 2(PACKET LOSS)
Router 2 (Linksys)
Parallel
Connections
2
5
2
5
2
5
2
5

80.00%

Fig. 13. Packet Loss Results Router 2 (Table IX)

Furthermore, Table VIII and IX expose the average jitter and percentage of lost packets, which corresponds to Router 2 performance indicators.

Router 2 (Linksys)
Parallel
Connections
2
5
2
5
2
5
2
5

5.0 UDP 2 20

60.00%

5.0 UDP 5 150 5.0 UDP 2 150 2.4 UDP 5 150 2.4 UDP 2 150

5.0 TCP 5 150 5.0 TCP 2 150 2.4 TCP 5 150 2.4 TCP 2 150

TABLE VIII.

40.00%

VI. ANALYSIS

Distance (ft)

Packet Loss

150
150
150
150
20
20
20
20

1.91%
5.46%
67.75%
78.70%
0.45%
0.67%
40.25%
90.33%

This final section explores the core of the project: the comparison (benchmark) of both wireless routers. In this manner, from the exposed results of both routers it is possible to calculate the 90% Confidence Interval of each set of measurements (32). Further, it is possible to determine a better performance of one over the other in each individual combination (e.g., which router has better throughput in this conditions: TCP connection on the 5.0 GHz band with 2 parallel streams at a 100 feet distance).
Therefore, for the first response variable (throughput over parallel TCP streams) we have:
TABLE X.

COMPARISON THROUGHPUT (150 FEET)
Confidence
Interval

Parallel
Connections

Average
Throughput
(Mbits/sec)

Min

Max

TCP

2

9.6736

9.341

10.0057

2.4

TCP

2

5.3072

4.799

5.8154

2.4

TCP

5

5.7620

4.646

6.8784

2

2.4

TCP

5

6.3169

5.913

6.7205

1

5.0

TCP

2

39.6083

38.022

41.1941

2

5.0

TCP

2

5.5941

5.149

6.0395

1

5.0

TCP

5

36.7984

35.137

38.4598

2

5.0

TCP

5

5.7859

5.361

6.2109

Frequency
(MHz)

Protocol

1

2.4

2
1

Router

0.00

2.00

4.00

5.0 UDP 5 20

6.00

8.00

5.0 UDP 2 20

10.00
2.4 UDP 5 20

12.00

14.00

2.4 UDP 2 20

5.0 UDP 5 150 5.0 UDP 2 150 2.4 UDP 5 150 2.4 UDP 2 150

Fig. 12. Jitter Results Router 2 (Table VIII)

16.00

Comparison
Router 1 Better
Throughput
It cannot be determined Router 1 Better
Throughput
Router 1 Better
Throughput

TABLE XI.

COMPARISON THROUGHPUT (20 FEET)

Router

Frequency
(MHz)

Protocol

Parallel
Connections

Average
Throughput
(Mbits/sec)

1

2.4

TCP

2

42.8203

2

2.4

TCP

2

8.9916

8.8639

9.1193

1

2.4

TCP

5

44.6169

44.2718

44.9620

2

2.4

TCP

5

8.9916

6.0905

7.1037

1

5.0

TCP

2

100.1848

96.3289

104.0407

2

5.0

TCP

2

6.9441

6.5781

7.3101

1

5.0

TCP

5

103.8958

102.9048

104.8869

2

5.0

TCP

5

7.1983

6.8187

7.5779

Confidence Interval
Min

Max

41.9171

43.7235

Comparison

2.4 GHz, TCP, 5 parallel streams and 150 feet. In this case, the
Confidence Intervals overlap; thus it is not possible to conclude which router has a better performance.

Router 1 Better
Throughput
Router 1 Better
Throughput
Router 1 Better
Throughput

For the second tested response variable (jitter over parallel
UDP connections) we have the following:
TABLE XII.

Router 1 Better
Throughput

COMPARISON JITTER (150 FEET)

Router

34

Parallel
Connections

Jitter
(ms)

2.4
2.4
2.4

UDP
UDP
UDP

2
2
5

2
1
2
1
2

39

Protocol

1
2
1

44

Frequency
(MHz)

2.4
5.0
5.0
5.0
5.0

UDP
UDP
UDP
UDP
UDP

5
2
2
5
5

TABLE XIII.

29

Min

Max

1.3685
4.5386
6.2372

1.3328
3.8683
5.9377

1.4042
5.2088
6.5367

8.7997
0.5872
5.8742
1.8602
15.0149

8.3099
0.5003
4.8841
1.6699
14.0068

9.2895
0.6740
6.8642
2.0504
16.0229

14

Frequency
(MHz)

Protocol

Parallel
Connections

Jitter
(ms)

1
2
1
2
1
2
1
2

19

Percentage
Lost
Packets

Comparison

0.16%
1.91%
47.37%

Router 2 More
Jitter
Router 2 More
Jitter

5.46%
2.26%
67.75%
9.02%
78.70%

Router 2 More
Jitter
Router 2 More
Jitter

COMPARISON JITTER (20 FEET)

Router

24

Confidence Interval

2.4
2.4
2.4
2.4
5.0
5.0
5.0
5.0

UDP
UDP
UDP
UDP
UDP
UDP
UDP
UDP

2
2
5
5
2
2
5
5

1.4564
1.7606
4.2698
4.0174
0.3693
4.2845
0.5581
5.3822

Confidence
Interval
Min
Max
1.3152
1.5976
1.6450
1.8763
3.7944
4.7451
3.8129
4.2218
0.3619
0.3766
3.5112
5.0577
0.5507
0.5655
5.0196
5.7448

Comparison

Percentage
Lost
Packets

Router 2 More
Jitter
Router 2 More
Jitter
Router 2 More
Jitter
Router 2 More
Jitter

0.00%
0.45%
0.00%
0.67%
30.43%
40.25%
47.55%
90.33%

9

18.00
4
0

1

2

3

R1/2.4/TCP/2
R2/2.4/TCP/5
R1/5.0/TCP/5

R2/2.4TCP/2
R1/5.0/TCP/2
R2/5.0/TCP/5

4

5

R1/2.4/TCP/5
R2/5.0/TCP/2

16.00
14.00
12.00

Fig. 14. CI Comparison (Router 1 vs Router 2) (Throughput 150ft)
(Table X)

10.00
8.00
6.00

120.00

4.00
100.00

2.00
0.00
0

80.00

60.00

1
R1/2.4/UDP/2
R2/2.4/UDP/5
R1/5.0/UDP/5

2

3
R2/2.4/UDP/2
R1/5.0/UDP/2
R2/5.0/UDP/5

4
R1/2.4/UDP/5
R2/5.0/UDP/2

5

Fig. 16. CI Comparison (Router 1 vs Router 2) (Jitter 150ft)
(Table XII)

40.00

20.00

0.00
0

1
R1/2.4/TCP/2
R2/2.4/TCP/5
R1/5.0/TCP/5

2

3
R2/2.4/TCP/2
R1/5.0/TCP/2
R2/5.0/TCP/5

4
R1/2.4/TCP/5
R2/5.0/TCP/2

5

Fig. 15. CI Comparison (Router 1 vs Router 2) (Throughput 20ft)
(Table XI)

As shown in Tables X and XI in terms of throughput over the TCP connections configuration, for all set of measurements, R1 (Netgear) presents a better performance compared to R2 (Linksys). The only exception is in the case of

As shown in Tables XII and XIII in terms of variation in the end-to-end delay (jitter) of the parallel UDP connections, for all set of measurements, R1 (Netgear) presents a better performance compared to R2 (Linksys). In other words, R1 has better variation of the end-to-end delay, which is considered a better performance especially for application layer processes running real time audio or video.
In this manner, based on the gathered measurements that were previously exposed, we can grade the performance of the wireless routers. Therefore, we can conclude, with a 90% confidence interval, that R1 (Netgear) has a better performance that R2 (Linksys) in terms of throughput (TCP), jitter and packet loss (UDP).

7.00



6.00

4.00

3.00

2.00

1.00

0.00
0

1
R1/2.4/UDP/2
R2/2.4/UDP/5
R1/5.0/UDP/5

2

3
R2/2.4/UDP/2
R1/5.0/UDP/5
R2/5.0/UDP/5

4
R1/2.4/UDP/5
R2/5.0/UDP/2

5

Fig. 17. CI Comparison (Router 1 vs Router 2) (Jitter 20ft)



VII. CONCLUSIONS



For the throughput response variable in TCP parallel connections, for all set of measurements,
Router 1 presents a better performance compared to Router 2. The only exception is in the case of
2.4 GHz, TCP, 5 parallel streams and 150 feet. In this case, the Confidence Intervals overlap; thus it is not possible to conclude which router has a better performance. o In the same light, for the second response variable, variation in the end-to-end delay (jitter) of the parallel UDP connections, for all set of measurements, Router 1 presents a better performance compared to Router 2. In other words, R1 has less variation of the end-to-end delay. Thus, we can conclude that the tested router Netgear
WNDR4300 has a better performance than the router
Linksys WRT600N. o 5.00



In addition, using Confidence Intervals (90%), this project explored the differences in performance of Router 1
(Netgear) and Router 2 (Linksys). Further, we have:

The proposed network performance study was successfully executed. Further, real measurements were taken, processed and analyzed in order to the comparison between equipment of different vendors.
The selected response variables (throughput, jitter and packet loss) were processed and analyzed from the original data, which was obtained through an iperf client/server architecture. Moreover, we observed the following circumstances: o

o

For both wireless routers in terms of throughput, we can observe that at less distance higher throughput. In the same way the 5.0 GHz frequency presents a higher throughput rate than the 2.4 band.
For the UDP measurements in the two routers, the amount of jitter and number of lost packets is directly impacted by the number of parallel streams and the distance. Thus, more streams and distance are translated in a bigger difference in the end-to-end delay (jitter)

REFERENCES
[1]
[2]
[3]

[4]
[5]
[6]

[7]

[8]

R. Golani, S. Kandalan, “WiFi Traffic Analysis Project Report”, Stony
Brook University”, New York.
Luxul Technology, “Use Iperf to Measure Wi-Fi Network Data Rates:
Data Rates vs. Signal Strength”, Lindon, UT.
J. Griffit, “How to Guide: Jperf and Iperf”, Wireless LAN Professionals, retrieved from: http://wirelesslanprofessionals.com/wpcontent/uploads/2011/02/How-to-Guide-on-JPerf-and-IPerf.pdf Iperf Support, “The Network Bandwidth Measurement Tool”, obtained from: http://iperf.fr
Netgear Support, “Manuals and Technical Information”, retrieved from: http://www.netgear.com/home/ Linksys Support, “Manuals and Technical Information”, obtained from: http://www.linksys.com/us/?gclid=Cj0KEQiA4qSzBRCq1iLhZ6Vsc0BEiQA1qt-zoQNczouEqTJHJnt2WPCMmLlnQDjVjZ02PSdmwG6RgaAkxR8P8HAQ Paessler AG, “Monitoring Tools Stay on Top of Your Network Using
Paessler’s
Monitoring
Tools”,
retrieved from: http://www.paessler.com/monitoring_tools
Kolahi, Samad S., et al. "Performance monitoring of various network traffic generators." Computer Modelling and Simulation (UKSim), 2011
UkSim 13th International Conference on. IEEE, 2011.…...

Similar Documents

Free Essay

Routers

...By: Dick Gilbert Routers A router is a network layer device responsible for routing traffic between subnetwork that make up a data internetwork . A subnetwork is defined as a collection of network resources that can be reached without needing to go through a router. A internetwork a collection of two or more subnetworks connected using routers. The role of a router is pretty simple. It receives packets transmitted by end stations and routes them through the internetwork to the right subnetwork. Routers can also perform more advanced functions if the network layer protocol supports those functions. Most of them have the ability to fragment packets. Routers also provide some congestion control capablilties .Routers can be fairly sophisticated packet filters. Some people sometimes mistake routing for bridging. Routing is also a network layer function. Routers operate on packets and choices based on the network layer address. Bridging is a data link layer function. Bridges operate on frames and make filtering and forwarding choices based on the media access control address. Packets ride within frames and are the transmission unit used to get a message from one end station to another anywhere in the internetwork. If two end stations are in the same subnetwork , the packet will be placed in a frame by the transmitter and sent across the subnetwork to......

Words: 369 - Pages: 2

Free Essay

Cisco Routers

...Bosack found a way to connect the two local area networks where he and his wife worked, 500 yards across campus. ! They initially tried to sell the inter networking technology that Bosack had developed to computer companies, but none were interested. With no buyers they decided top start their own company, Cisco Systems. ! Cisco’s primary product from the beginning was the inter networking router; a hardware device incorporating software that automatically selects the most effective route for data to flow between networks. Cisco thus became the first company to provide a multi-protocol router when it shipped its first product in 1986. One year later after initial start, Cisco was selling $250,000 worth of routers per month. Sales for the fiscal year ending July 1987 were 1.5 million, and the company had only eight employees at the time. ! In the beginning, Cisco marketed its routers to universities, research centers, the aerospace industry, and government facilities by contacting computer scientists and engineers via ARPANET. In 1988 the company began to target its internetworking routers to mainstream corporations with offices in multiple locations that used different networks. ! Cisco had a high rate of sales growth in the early years, however they were increasing in need of cash if they were going to continue to grow. Bosack and Lerner were forced to turn to a venture capitalist. They turned to Donald T. Valentine of Sequoia Capital, however, he required them to......

Words: 1459 - Pages: 6

Premium Essay

Routers

...Routers Term Paper INTRODUCTION Router is a hardware device designed to take incoming packets, analyzing the packets and then directing them to the appropriate locations, moving the packets to another network, converting the packets to be moved across a different network interface, dropping the packets, or performing any other number of other actions. In the picture to the down, is a Linksys BEFSR11 router and is what most home routers look like. [pic] A router has a lot more capabilities than other network devices such as a hub or a switch that are only able to perform basic network functions. For example, a hub is often used to transfer data between computers or network devices, but does not analyze or do anything with the data it is transferring. Routers however can analyze the data being sent over a network, change how it is packaged and send it to another network or over a different network. For example, routers are commonly used in home networks to share a single Internet connection with multiple computers. [pic] In the above example of a home network there are two different examples of a router, the router and the wireless router. As can be seen in the example the router is what allows all the computers and other network devices access the Internet. Below are some additional examples of different types of routers used in a large network. Brouter Short for Bridge Router, a "brouter" is a......

Words: 569 - Pages: 3

Free Essay

Benchmarking

...Benchmarking Training Article Benchmarking for competitive advantage Introduction There has been a progressive increase in the topic of benchmarking – but for all that there is still a great deal of ignorance about what it actually is. It has taken its place as a management buzzword along with BPR, TQM, Change, EVA and many others – but its true nature is poorly understood. Some see it as stealing (or ‘borrowing’) ideas; others as a mechanism for comparison with a competitor; whilst others view it as a form of industrial espionage. In fact it is all of these and none of these at the same time, but instead involves understanding strategic gaps; cooperation; hard work; a willingness to question and where necessary to change fundamental precepts (sacred cows) and also - giving. This paper will address several issues including: what benchmarking is; when it should be used; and the benefits from using it; how to approach a benchmarking initiative; and how to optimise a benchmarking project. What is benchmarking? The term benchmark comes from surveying where it was used to denote a notch or mark representing a given altitude and against which other heights could be calibrated or ‘benchmarked’, since when it has come to mean any standard against which something is compared; and some of the leading exponents in business include Xerox and GE. In business terms there are numerous definitions of benchmarking, but essentially it involves learning, sharing information and......

Words: 4668 - Pages: 19

Free Essay

Benchmarking

...Benchmarking El benchmarking es un anglicismo que, en las ciencias de la administración de empresas, puede definirse como un proceso sistemático y continuo para evaluar comparativamente los productos, servicios y procesos de trabajo en organizaciones. Consiste en tomar "comparadores" o benchmarks a aquellos productos, servicios y procesos de trabajo que pertenezcan a organizaciones que evidencien las mejores prácticas sobre el área de interés, con el propósito de transferir el conocimiento de las mejores prácticas y su aplicación. En economía, toma su acepción para la Regulación: benchmarking es una herramienta destinada a lograr comportamientos competitivos (eficientes) en la oferta de los mercados monopolísticos, consistente en la comparación del desempeño de las empresas, a través de la métrica por variables, indicadores y coeficientes. En la práctica, se utilizan diversos mecanismos de incentivos al comportamiento eficiente, como la publicidad de los resultados de las comparaciones (nadie quiere ser el peor) o con la utilización de mecanismos que transforman esos resultados comparativos en premios o castigos sobre los ingresos del empresario. 1. Reseña histórica del benchmarking En 1982, en Rochester, durante una reunión de la Corporación Xerox donde se trato el aspecto organizacional de esta, con respecto a sus competidores, se utilizó por primera vez la palabra Benchmarking Competitivo y cuyo sistema impresionó por la manera en que se recopilo la información. Allí......

Words: 7110 - Pages: 29

Premium Essay

Benchmarking

...Benchmarking is a measurement of the quality of an organization’s policies, products, programs, strategies, etc., and their comparison with standard measurement, or similar measurement of its peers. The objectives of benchmarking are; 1 To determine what and where improvements are called for. 2 To analyze how other organizations achieve their high performance levels. 3 To use this information to improve performance Benchmarking is also called; the process of identifying best practices in relation to both products and services. I work for a non-profit organization where we use strategic benchmarking to indicate where the organization needs overall best performance by examining the long terms strategies and general approaches that have enable high performances to success. When you operate on grants, funding sources require high performance and accountability in order to renew their awards-contract each year. The organizations are reimbursed for service provided on performance-based. So there is a constant challenge to stay in business. In performance there is also competitive benchmarking to deal with other organization providing the same services like the public schools, the nursing homes, the hospitals, and the government funded universities as well. Non-profit organizations involve in internal benchmarking for measurement of effective use of available resources, performance evaluation within departments, employees satisfaction, and wages comparison in HR, employee welfare.......

Words: 252 - Pages: 2

Free Essay

Benchmarking

...-BENCHMARKING - 2014 BENCHMARKING DIANA C. HURTADO MUÑETON LORENA PULIDO BENAVIDES DANIEL JIMENEZ GOMEZ DOCENTE: EFRAIN OCAMPO UNIVERSIDAD AUTONOMA DE MANIZALES NEGOCIOS INTERNACIONALES COSTOS LOGISTICOS Y BSC 1 -BENCHMARKING DEFINICIÓN El benchmarking es una forma de determinar qué tan bien se desempeña una empresa (o una unidad de esta), comparado otras empresas (o unidades). Esta ubica el desempeño de la empresa en un contexto amplio, y permite determinar las mejores prácticas.  “Benchmarking es el proceso continuo de medir productos, servicios y prácticas contra los competidores más duros o aquellas compañías reconocidas como líderes en la industria”. David T. Kearns, director general de Xerox Corporation. La continuidad en el benchmarking no sólo significa un proceso que se hace una vez y se olvida, sino que es un proceso continuo y constante. Otro aspecto es el de la medición, ya que esta está implicada en el proceso de benchmarking, pues se tienen que medir los procesos propios y los de otras empresas para poder compararlos. Del mismo modo este proceso se puede aplicar a todos las facetas del negocio. Y finalmente este se debe dirigir hacia aquellas empresas y funciones de negocios dentro de las empresas que son reconocidas como las mejores o como los líderes de la industria.  “Una marca del agrimensor de una posición previamente determinada y que se usa como punto de referencia, un estándar mediante el cual se puede medir o juzgar algo”....

Words: 2416 - Pages: 10

Premium Essay

Benchmarking

...Company Benchmarking for Competitive Advantage Terry Pilcher Assistant Director – Department of Trade and Industry UK Before undertaking a benchmarking study a company must be aware of the different types of benchmarking that exist. One of the major reasons for misunderstanding is the tendency to call a wide variety of different activities Benchmarking. The most simplistic form of benchmarking could be one person talking to another and deciding if they have similar areas of interest. At this level people undertakes benchmarking every day. Before listing the different types of benchmarking that exist it might be useful to present some definitions of benchmarking to set the scene. There are many definitions around but it is not so much the differences that matter but the similarities. This can be best illustrated using the following definitions: ‘A continuous systematic process for evaluating the products, services and work of organisations that are recognised as representing best practices for the purpose of organisational improvement’ (Spendolini, 1992) ‘Benchmarking is an external focus on internal activities, functions, or operations in order to achieve continuous improvement’ (McNair and Leibfried, 1992) ‘A disciplined process that begins with a thorough search to identify best practice organisations, continues with the careful study of one’s own practices and performance, progresses through systematic site visits and interviews, and concludes with an analysis of......

Words: 1707 - Pages: 7

Premium Essay

Benchmarking

...The definition of benchmarking is “the continual process of measuring products, services and practices against the toughest competitors or those companies recognized as industry leaders” (Wheelen & Hunger, 2010). Benchmarking focuses on bettering a company by openly learning how other companies operate and so competing companies can better their business operations. There are usually six steps involved with benchmarking and they are known to be used within companies that are already well managed, instead of focusing on companies that are in poor standings. To me, benchmarking is a positive program. I do not believe it is just another fad. “Benchmarking can help with increase sales, expand goal setting, and lift employees drive” (Wheelen & Hunger, 2010). To look into benchmarking, “the average cost of a benchmarking study is $100,000.00 and involves 30 weeks of effort” (Wheelen & Hunger, 2010). By reading about benchmarking, it seems as if the study would be well worth the money and effort (time put into bettering a company) because the outcome would be on top of what was spent. If my company was already doing well and I took on a benchmarking study and come out on top of what I spent, the time, effort and money would not be a waste. In fact, it would be a positive resource brought into my company. I do believe it is a valuable tool now and will continue to be in the future. References             Wheelen, T. L., & Hunger, D. J. (2010). Concepts in......

Words: 264 - Pages: 2

Free Essay

Benchmarking

...The Practice Of Benchmarking And Its Advantages And Disadvantages Introduction Over the years, there has been a tremendous development in benchmarking techniques, in terms of both benchmarking practices and method, which are widely used in different organization to achieve different goals in the developing countries. In this assignment I will discuss the definition of benchmarking and then the use of benchmarking. After that I will discuss the advantage and the disadvantage of benchmarking according the researches and the case studies of benchmarking. Then, I will discuss and analyze the practice of benchmarking to explain why managers find that benchmarking did not live up to their expectation. Finally, the conclusion will have summary of the assignment question. Background of Benchmarking 2.1 Definition of Benchmarking Benchmarking, as the definition of the CEO of Xerox -D.T. Kearns- is the continuous process of measuring the products, the services, and practices against the best competitors or the leader in their industry (Kolarik, 1995). Freytag and Hollensen (2001) states that ‘benchmarking involves measurement of business performance against the best and makes continuous effort in reviewing process practice and method’. This essay is an example of a student's work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay......

Words: 2696 - Pages: 11

Premium Essay

Benchmarking

...The basic idea behind benchmarking is by doing a comparison. An organization finds another organization that is similar and compares them. By studying the achievements of such results; make plans for improving the business performance; implement plans monitor and evaluate the results. By doing the comparison it allows the organization to make the necessary changes, but it is not a quick fix. Benchmarking is vital, by using this practice it can identify costs and provide efficient information that can implement best practice by becoming a central instrument for improvement and the performance of an organization; under the right conditions comparison can be a key driver of performance (OECD, International benchmarking, p. 1). Through benchmarking, many organizations realize that more can be achieved through collaboration between leaders in an industry. There are three primary types of benchmarking used today; first is process benchmarking, which is done at the lower levels of the organization. It focuses on the day to day activities and operations to include anything from customer complaints to the billing process. The changes made on this level improve performance quickly. Second there is performance benchmarking, focuses on competitiveness by looking at where the company’s products and services are in relation to other competitors referring to things such as speed, quality and reliability. Third there is strategic benchmarking, focusing on top management. Discussions are......

Words: 644 - Pages: 3

Premium Essay

Benchmarking

...The Importance of Benchmarking in Improving Operational Performance Benchmarking is a key tool used in all sectors in the industrial world. It is an accepted standard to which actual performance is compared, evaluated and, if needed, improved. In the current economy there is an increase in competition therefore companies have to always be ahead of their competition by having a high standard of quality maintained, the companies need to be constantly updated with the newest technology, and at the same time have lesser costs than their competition. “Benchmarking helps to improve the company and helps the company with innovate. It is the process of identifying, understanding, and adapting outstanding practices from organizations anywhere in the world to help an organization improve its performance. It is an activity that looks outward to find best practice and high performance and then measures actual business operations against those goals” (Kumar et al., 2006). This definition is the most evolved definition of benchmarking; benchmarking went through four stages of evolution. It is heavily used in the retail industry to help improve the quality of a retail business, which then leads to higher profits. There are three points that need to be covered to have a successful level of benchmarking: the business must focus around its customers and employees and have continuous improvements; the business has to be flexible to change and has to be able to share and accept all......

Words: 1493 - Pages: 6

Premium Essay

Benchmarking

...TITLE: Benchmarking: A general review about the classification, process and pitfall of the model LECTURER: PROF. MADYA DR. RIZAL BIN RAZALLI NAME: ONG YUH KIANG MATRIC NO: 227806 TITLE: Benchmarking: A general review about the classification, process and pitfall of the model LECTURER: PROF. MADYA DR. RIZAL BIN RAZALLI | | | | | | Summary of selected articles Article 1 | Benchmarking: a general reading for management practitioners | Author | Sik, W.F., Cheng, E.W.L., & Ho, D.C.K. | Year | 1998 | Setting/ Place/ Sample | Management Decision, Vol. 36 Iss 6 pp. 407 - 418 | Objectives | Introduce the basics of benchmarking to management practitioners | Analysis | Classification of benchmarking, benchmarking process model, components in a benchmarking project, hints for organization to conduct first time benchmarking project. | Findings | Identify the sources of best practices as a result of assisting in promoting management perspectives and facilitating the formation of various specific goals. | Article 2 | Benchmarking the benchmarking models | Author | Anand, G., & Kodali, Rambabu | Year | 2008 | Setting/ Place/ Sample | Benchmarking: An International Journal, Vol. 15 Iss 3 pp. 257 - 291 | Objectives | To classify the fundamental benchmarking model and the unique benchmarking model, propose a universal benchmarking model which can be applied for all types of benchmarking | Analysis | Benchmarking the benchmarking......

Words: 3835 - Pages: 16

Free Essay

Benchmarking

...WHAT IS BENCHMARKING? http://tqmgroups.blogspot.com/p/benchmarking.html “Benchmarking is simply the process of measuring the performance of one's company against the best in the same or another industry. Benchmarking is not a complex concept but it should not be taken too lightly. Benchmarking is basically learning from others. It is using the knowledge and the experience of others to improve the organization. It is analyzing the performance and noting the strengths and weaknesses of the organization and assessing what must be done to improve. REASONS FOR BENCHMARKING There are several reasons that benchmarking is becoming more commonly used in industry; •         Benchmarking is a more efficient way to make improvements. Managers can eliminate trial and error process improvements. Practicing benchmarking focuses on tailoring existing processes to fit within the organization. •         Benchmarking speeds up organization’s ability to make improvements. •         Compare business practices with those of world class organizations •         Challenge current practices and processes •         Create improved goals and practices for the organization •         Change the perspective of executives and managers. OBJECTIVES OF BENCHMARKING ϖ  Becoming competitive ϖ  Improving industry best practices ϖ  Defining customer requirement ϖ  Establishing effective goals and objectives ϖ  Developing the measures of......

Words: 2078 - Pages: 9

Premium Essay

Wireless Router

...Plus Wireless Router While exploring my text book for a topic for my term paper I chose a subject that I have recently been involved with. Several of us have configured and installed a basic home wireless network. In many homes today it is very popular to have a wireless router installed in your home office creating a connection of convenience to the internet. Installing a wireless router is relatively easy. At the beginning of this semester I had to replace my old wireless router due to an electrical surge through my Ethernet port from my modem. While looking into the purchase of a new wireless router I came across the Cisco Valet Plus Wireless Router. The purpose of this paper is to provide the manufactures specification for the Cisco Valet Plus Wireless Router, the function, the appearance, the operation, and the pros and cons of this device. I chose the Cisco Valet Plus Router Based on the specifications listed on the products exterior package. What caught my attention were the Data Link Protocol; Ethernet, Gigabit Ethernet, IEEE 802.11b, IEEE 802.11n, IEEE802.11g, Fast Ethernet, and the data transfer rate 300 Mbps. In today’s day and age it seems the speed and efficiency is very necessary and takes priority in our daily lives. The frequency band is 2.4 GHz. A feature that I have never seen offered in a wireless router is the Anti-Malware protection, Wi-Fi Protected Setup (WPS), URL Filtering, Domain blocking, and Parental control. This is the first wireless......

Words: 1198 - Pages: 5