ERRATA - [Your Contributions] Last Modified 5/2/2001
Latest Printing
***Page 71, Figure
3.18: The "DATA" lines
should go up and to the right,
not down, to be consistent
with step 2 on Page 70.
***Page 73, top line: "0xC0" should be
"0xc0" (lower case c) for
consistency.
***Page 74, 3.2 should have
a bold N(n), not italic; 3.6 should have a bold
U(n), not italic; in 3.3 and
3.4, d-sub-n should be d-sub-n(i).
Putting the = in 3.3 instead
of in 3.4 would mean no change unless there is
an improvement; as written,
the path is changed to go through more nodes
without an improvement in
"time."
***Page 75, top: The proof assumes that j is a
previously considered node
without giving a
justification.
***Page 76, middle
paragraph: The text says a line is
made thick, when it
is actually made green.
***Page 79, paragraph after
"Description" L(i,j)=L(j,i) need not be one,
just some number less than
infinity.
***Page 109, first
paragraph: "A chooses some
random time m*rho..." should
be "2*m*rho". A
and B are reversed in the second part of the paragraph. The
last ">rho,"
should be ">=rho" to allow for m-n=1.
***Page 113, second
paragraph: Should the reference to
(4.7) be (4.5)? In
4.6, are we to conclude that
5 * 256 is approximated as 1000?
The sentence
ending "... 27
percent." is stylistically inconsistent with Page 110 (par.
second from bottom) and
elsewhere, where the % symbol is used, not the word.
***Page 116, 4.10.2, first
line: Rho is not bigger than
2TTRT.
***Page 118, second
equation: It says that A(n) + B(n)
+ rho + A1prime +
B1prime is <= TTRT. The sums are collapsed using
definitions 4.15 and 4.16.
Can this be derived from the
hypotheses 4.17, A(n) + rho <= TTRT?
***Page 121, equation 4.24:
It should be: S = G x P(0,1) =
G*exp(-G).
There are two errors: the G is missing in the middle part
(inconsistent
with 4.22) and the desired
probability is of ZERO new transmissions, GIVEN
that one is in progress i.e.
a transmission is successful if there are no
collisions.
***Page 123. The first word "figure" in a
line in the long paragraph has no
number and seems to refer to
a nonexistent figure.
***Page 137. (up to 105!) could be read as
105-factorial.
***Page 235, 5th
paragraph: ADLS should be ADSL.
***Page 285, Problem 8: Unless it is intended that the answer
is that it is
impossible to send any
traffic down path two (p=1), the delay of 0.10
microseconds is too
low. Equation 9.6
T = 1 / (10Mbs - lambda)
gives a delay of approximately 0.10 * (1 +
lambda*10^-6) microseconds.
This is for the first stage
of the second path alone. Of
course, sending
all traffic through the second path has a delay of 0.1111... plus packet
retransmission time at the second stage of path two.
***Page 290, the third
equation, the one below A.6: The
P{Y>=0} should be
P{Y>0}. The former is the probability of a
non-negative integer, which is
one, since it over the whole probability space. The same applies to the numerator.
***Page 291, Equation
A.12: The n=0 in the middle sum
should be n=1. As it
is, it has (0-1)! as the
first term, and (-1)! implies that the Gamma
function is defined at 0, which means that you can integrate 1/x at x=0.
Thanks to Paul Donis and Dr. S. C. Chang at George Mason University.
Earlier Printing of Book:
p2 figure 1.2(a)
A link is missing between the upper right hand corner node and the lower
left hand corner node.
p. 18, 2.1.2, par 3 line 1: Change "in by" to "by"
p. 23: Description of packet from router W to computer D should read
[w,d|S,D|data] rather than [r,d|S,D|data] (last sentence of third paragraph)
p24, first paragraph
"We examine [four] three increasingly sophisticated procedures"
p. 26: 2.2.4, par 3, line 4: Change "trough" to "through"
p. 27, 2.3.2. Thus the virtual circuit from B to D uses the number 1 from B to
S2 and the
number 2 from S2 to S3 and from S3 to D.
p. 29, Figure 2.10: Transmit only if transmission does not make buffer overflow.
p 30, 2.4 Network Architecture, 4th paragraph, 4th sentence
Should put a "and" between links and servers.
p 31, 32
For figure 2.11 we should have
a brief description in the diagram of the lower three layers.
p. 33, Figure 2.13: You may want to add SNMP on top UDP.
p 34, figure 2.15
In the text above and in the text of the diagram, the wording is not
consistent. We use "data" in the text, but "user" in the
diagram.
p. 41, Statistical Multiplexing, pp2: ``most transactions take between a
fraction of a second and some take up
to a few minutes.'' You should remove the word ``between'' from the sentence.
p. 45. Figure 2.29: Remove the three "1"s on the links.
p46
IETF stands for Internet Engineering Task Force
p.51, Figure 3.6: You may want to add SNMP on top UDP.
p. 55 fourth line. Change "the a" to "a"
p57
IETF stands for Internet Engineering Task Force
p. 64. Figure 3.14: black should be blue; green should be black
(according to text).
p. 65, about the middle of first paragraph: "Note that link CF
produces a smaller estimate (6)..." It must be (5)
p. 73, 3.7.1, par 2: Add "P_1(a)={a}"
p. 74: Remove equation 3.4
p. 74. 3.7.1, par 3 line 5: change "step (3.2)" to "step (3.3)"
p. 76 equation 3.8: Change "L(i,j)" to "L(k,j)"
p. 79 3.8.1, par 3, line 6: Change "reach" to "react"
p. 85 Figure 3.24: Remove "D" from "D,E" network
p. 104, Figure 4.11: GMSK instead of GSMK
p. 105, par 2 line 4: Change "A, C, F" to "A, C, E"
p. 111, 4.8.4 par 3: change 0.176 in equation to 0.176(5)
change 6.3 Mbps to 5.3 Mbps (twice)
p. 112: the value of tau = 1 microsecond must be changed to 80 microseconds or one needs to change the packet length to 1000 bits instead of bytes. In either case, the value of beta must be changed accordingly.
p. 113, 4.8.6: change 216 to 256 four times
p. 135 Figure 5.8. Add "b,1-c,2" to middle routing table.
p. 139 fourth last line. Remove sentence "In addition ... output links."
p. 162 line 7. Change "SRP" to "SWP".
p. 162 Change to
theta=00A, L={01B}, R={00A,01B}, V={01B}
theta=01B, L={11B}, R={00A, 01B, 11B}, V={11B}
theta=11B, L={10A}, R={00A,01B,11B,10A}, V={10A}
theta=10A, L={00A}, R={00A,01B,11B,10A}, V=0.
p. 163 6.3.3, par 2, line 10, change p to 1-p (the prob. of X=T+Y)
p. 164 6.4.2 line 7, change "Internet" to "the Internet".
p. 165 line 5. Define PTA as "packet-to-acknowledgement system"
p. 167 6.5.2 after eq (6.3), change "T>=W" to "T>=W * tau"
p. 190 last line of 6.11.0 Change "mechanism" to "mechanisms"
p. 191 third line of 6.11.3 Change "packets" to
"packet".
p. 192, 6.11.3, last para, line 6: change "wait" to "waits"
p. 192, 6.11.3, last para, line 8: change "figure 3" to
"figure 6.27"
p. 193. Tahoe incorporated slow start, congestion avoidance, and fast
retransmit. Reno added fast recovery.
p. 193 Figure 6.28 Line 6 (E5), the last packet in transit is L+K+2 and
the last received packet is L+K+1. Line 7, the last packet in transit is L+K+3
and the last received packet is L+K+2.
p. 195 line 2: italicize x
p. 195 line 7: change "connection" to "connections"
p. 195 6.11.6, par 2, line 1: change "for" to "form"
p. 221. All references to QPSK should be to 8PSK
p. 235, 7.10, next to last line: change "ADLS" to "ADSL"
p. 236 Problem 5. Change "Exlain" to "Explain"
p. 241 8.2 par 1, line 4. Change "private key" to "public key"
p. 244 8.3.1 MAC, line 2, change "E(H(P) o K)' to "E(H(P);K)"
p. 245 8.3.2 DH Exchange, line 1 should read "This strategy is named..."
p. 246 8.3.2 2nd last line should read "It then suffices..."
p. 253 fourth last line. Change "reaches a leave" to "reaches a leaf"
p. 261 8.7.1 line 2: the one-time pade is explained in 8.2.1
p. 262 8.8 line 1: only one code is explained
p. 269: The reference on information theory is Shannon (1948), not McGraw
(1997), obviously.
Pg. 311, near bottom of page, Li is the average *number* in the queue, not
the average delay.
Back Cover: ADSL [instead of ADS]
Thanks to Jeff Caruthers (Boston University), Teri Centner (George Mason University), Scott Jordan and John Kristoff (Northwestern University), Mike Sheppard (University of New Mexico), Hector Simosa (Department of EE, University of Massachusetts, Lowell Campus), Xiren Cao (UST, Hong Kong).