Introduction to the Internet

The internet has revolutionized technology since its rise in popularity. It has

improved peoples’ lives and brought the world closer together. In this report we’ll give

an overview of internet control, modeling of a FSK modem, analysis of packet loss in the

modem, and the improvement of transfer speed of the modem will be discussed.

Many people have a misconception that the internet is the World Wide Web

(websites like [Link] and eBay). However, these are only a part of the internet. By

definition, the internet is a worldwide, publicly accessible network of interconnected

computer networks that transmits data by packet switching using the standard Internet

Protocol (IP). It is actually consists of many different aspects. Generally, the internet has

two parts- the physical and non-physical net. The physical net has many important

components that include core network, fiber optic lines, backbone, router, modem, Local

Area Network (LAN) and computers. The core network is where everyone gets his or her

“internet” from. From the bottom of the network, consumers access websites from local

Internet Service Provider (ISP). The local ISP then gets their access from a regional ISP

such as AOL or Comcast. Contrary to popular belief, a regional ISP is still not the

ultimate service provider; they get their access from Network Service Providers (NSP),

such as IBM or UUNet, as well as Metropolitan Area Exchanges (MAEs). A backbone is

also a part of the core network consisting of many different networks providing

connection facilities for their clients. Fiber optic lines are what physically connect the

internet globally. In a local network, a router forwards data packets across the network

toward their destinations. What a router does is analogous to what a street sign does at an

intersection. At the local machine level, a modem is a device that modulates an analog

1
carrier signal to encode digital information to send across the network. It also

demodulates carrier signals, decoding the transmitted information. Our team is going

MatLab to simulate what a modem does. Lastly, computers (local machines) are needed

to connect end-users with the internet. The most notable parts of the non-physical net are

the Internet Protocol (IP), which communicates data across a packet-switched network,

routing table, which direct forwarded data by matching destination addresses to the

network paths used to reach them, router software and the World Wide Web, which

provides users with spaces to read, write and share information.

Most end-users want fast and reliable internet connections. However, many things

can cause the internet to have poor performance. Slow performance can be cause by

things like complex page assembly, poor HTML coding and outdated end-user machines.

Many users’ computers are outdated, which make them run very slowly. Improving

internet performance is something our team did extensive research on.

By the time of mid-term presentations, our team has found two main solutions:

NAT and IPV6. Network Address Translation (NAT) assigns the same network address

to multiple computers on the same network. It is very secure and provides reliability and

convenience for network administrators. More importantly, it saves IP addresses, since

there are only a limited number. Currently, the IP standard is IPv4. After many years of

research and testing, a new IP standard, IPv6, is being developed. This new standard fixes

address space problems, standardizes headers and simplifies routing tables. It is, overall,

a much better standard, but unfortunately, China is the only country that has fully adopted

IPv6. Once every country adopts this new standard, the internet should see a considerable

improvement in performance.

2
 As mentioned earlier, our team spent the second half of the semester developing a

way to simulate an FSK modem in Simulink. Our simulated modem has all the

functionalities of a regular modem and produces a signal that can be transmitted easily

and decoded to reproduce the original digital data.

Concept of Frequency Shift key

Frequency-shift keying (FSK) is a form of frequency modulation in which the

modulating signal shifts the output frequency between predetermined values. Multiple

frequency-shift keying (MFSK) is a variation of frequency-shift keying (FSK) that uses

more than two frequencies. In the following Simulink block diagram, the frequency shift

key in a modem used in the application of current internet control. The most fundamental

concept of the function of the modem system is it can convert the digital binary signal

from the computer to a sinusoid curve with varying frequencies for transmission along

the cable and back to binary signals to be input into the computer after it reaches its

destination.

3
 Figure 1. Different signals in a modem system.

Simulink Model

The Simulink model consists of a Bernoulli binary generator, M-FSK Modulator

Baseband block, AWGN Channel, M-FSK Demodulator Baseband block, relational

operator, spectrum scope, scope, integer delay and error rate calculator.

4
 Figure 2. Frequency-Shift key Simulink block diagram.

The Bernoulli Binary Generator block generates random binary numbers using a

Bernoulli distribution. The Bernoulli distribution with parameter p produces zero with

probability p and one with probability 1-p. The output is a binary signal, with random

combination of ‘1’s and ‘0’s.

The M-FSK Modulator Baseband block modulates using the M-ary Frequency

shift keying method. The output is a baseband representation of the modulated signal.

The M-ary number parameter, M, is the number of frequencies in the modulated signal.

The frequency separation parameter is the distance, in Hz, between successive

frequencies of the modulated signal. The input and output for this block are discrete-time

signals. Whether the input is an integer or a binary representation of an integer, the block

5
maps the integer 0 to the highest frequency and maps the integer M-1 to the lowest

frequency. In baseband simulation, the lowest frequency is the negative frequency with

the largest absolute value. The AWGN Channel block, from the Channels library, models

a channel using additive white Gaussian noise, with noise distributed over a bell shaped

distribution curve.

The M-FSK Demodulator Baseband block demodulates a signal that was

modulated using the M-ary Frequency shift keying method. The input is a baseband

representation of the modulated signal. The input and output for this block are discrete-

time signals. Whether the output is an integer or a binary representation of an integer, the

block maps the highest frequency to the integer 0 and maps the lowest frequency to the

integer M-1. In baseband simulation, the lowest frequency is the negative frequency with

the largest absolute value.

The Relational Operator block, from the Simulink Math library, compares the

transmitted signal, from the Bernoulli Random Binary Generator block, with the received

signal, from the M-FSK Demodulator baseband block. The block outputs a zero when the

two signals agree, and a one when they differ. If the result is TRUE, the output is 1; if

FALSE, it is 0.

The Scope block, from the Simulink Sinks library, displays the transmitted signal,

the received signal, and the output of the Relational Operator block. In the model

simulation, 3 data signals will be collected on the scope. The first one is the original

6
binary signal generated by the Bernoulli binary generator with an integer delay. The

second data signal is the output binary signal of the Demodulator Baseband block, which

has Gaussian white noise added to the input binary from the Bernoulli binary generator.

The last signal displayed will have the signal to identify any difference between the input

binary data and the output binary data of the model. The Error rate display block displays

the number of errors introduced by the channel noise.

Frequency Shift Keying

Frequency Shift Keying is a data signal converted into a specific frequency or

tone in order to transmit it over wire, cable, optical fiber or wireless media to a

destination point. FSK dates back to the early 1900s when this technique was discovered.

It was used to work alongside teleprinters to transmit message by radio (RTTY).

Although FSK is so old it still can be useful. With some modifications is can still work in

the world of computer and modems. In fact, the contributions of FSK are much more far

reaching. For example, the principle of FSK has laid the path to the development of other

similar techniques such as the Audio Frequency Shift Keying (AFSK) and Multiple

Frequency Shift Keying (MFSK) just to name a few.

Technically FSK has two classifications, the non-coherent and coherent FSK. In

non-coherent FSK, the instantaneous frequency is shifted between two discrete values

named mark and space frequency, respectively. On the other hand, in coherent Frequency

Shift Keying or binary FSK, there is no phase discontinuity in the output signal.

7
 In this digital era, the modulation of signals are carried out by a computer, which

converts the binary data to FSK signals for transmission, and in turn receives the

incoming FSK signals and converts it to corresponding digital low and high, the language

the computer understands best.

An Example of this can be seen in figure 3. The binary input signal is represented

by two frequencies. A zero has a different frequency then a one. By changing between

these two frequencies the computer on the other end of the line can distinguish and

convert/recover the original signal.

Figure 3. Binary signal conversion

Multi Frequency Shift Keying

MFSK is slightly different than FSK. MFSK works the same way as FSK, but

rather than two different frequencies, many more are used - from six to 64, for example.

8
MFSK16 uses 16 tones. Each tone transmitted is just like a Morse code dot, but they

follow each other without pause, on slightly different frequencies.

MFSK has some advantages above FSK, it is not only faster then FSK, but it is

also less sensitive to noise. MFSK can be improved with FEC (Forward Error Correct) to

make it even more reliable. FEC reduces the number of received errors. FEC takes up a

lot of bandwidth, but it makes up for it by making the system nearly perfect.

So FSK and MFSK are methods to convert an digital signal to an analog signal

and back. These methods have been used successful around the world for many years.

The future however might not be in MFSK. For the future we need something even more

advanced.

Results

Using the modem model that we created we were able to test to see the effects of

variation in line noise. We then began to vary the signal to noise ratio in the noise line in

a sweep from 1 to 10 db. A sweep was then performed for a higher mary number

changing the model from FSK to MFSK. An examination of the bit error rates caused by

this noise can be seen in Figure 4 below.

9
 Bit Error Rate
0.07

0.06
FSK
MFSK
Bit Error Rate (%)

0.05

0.04

0.03

0.02

0.01

0.00
0 2 4 6 8 10

SNR (db)

Figure 4: Bit Error Rate

From the chart one can see that the bit error rate of a SNR of 1 decibel, an incredibly

noisy signal was still comparatively small for both FSK and MFSK. The bit error rates

were both less then a fraction of a percent.

The bit error rate effect seemed negligible, however true Internet transfer occurs

via packet transfer. For the purpose of this analysis we used 1500 bytes for the packet

size. Although Ipv4 packets vary in size, 1500 bytes is a common value which also

corresponds to the standard packet size for Ipv6 packets. Individual packet loss can then

be calculated simply as the chance that a packet has no individual bit errors. Given

according to the equation (1-biterrorrate)^(#ofbits) which in this case is

1500bytes*8bits/byte. Figure 5 below of the resulting packet loss rates for the FSK and

10
MFSK cases.

Packet Loss

100.00
FSK
90.00
MFSK
Packet Drop Rate

80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0 1 2 3 4 5 6 7 8 9 10

SNR (dB)

Figure 5: Packet Error Rate

From the packet loss rates results we can see that there is a virtually 100% packet

loss rate above a certain noise threshold and that it rapidly transitions down to virtually

nil packet loss rate. This effect is due to the compound nature of packet loss, any bit error

rate when compounded 12000 times causes dramatic packet loss. The important result

from this section however is the difference in these noise thresholds. The FSK model cut

off was ~6dB while the MFSK was ~7.6dB.

11
Conclusions

From our results we can see that changing from FSK to MFSK requires a cleaner

signal transmission on the lines. Comparing this FSK modulation (an old phone line type

of modulation) with some industry numbers for QPSK (SNR above 12dB), 16bit QAM

(above 16db), and 64 bit QAM (above 24dB) we can see that a more advanced encryption

modulation technique requires a cleaner line signal. This effect is caused by the

increased modulation and demodulation resolution employed by the more advanced

techniques. This is the underlying principle behind Internet transfer to DSL and cable

modems. Traditional phone lines are too noisy to employ more advanced techniques, the

DSL and cable lines are employed to gain cleaner signals and hence increase through

rate. As an additional recommendation, it can be noted that with FSK there are discrete

frequencies of interest. It would be quite possible to add high and low pass filters to

reduce the signal noise in order to implement a more advanced technique.

12