Tetcos Engineering

TETCOS

Transmission Control Protocol

(TCP)
Comparison of TCP Congestion Control
Algorithms using NetSim™

@2017 Tetcos. This document is protected by copyright, all rights reserved

Tetcos Engineering

Table of Contents

1. Abstract………………………………………………………………………………………..…….. 4

2. Introduction to TCP…………………………………………………………………….........…........ 4

2.1. Problem of congestion…………………………………………………………………………. 4

3. TCP Variants…………………………………………………………………………………………. 5

3.1. Old Tahoe………………………………………………………………………………………... 5

3.1.1. Slow Start………………………………………………………………………………… 5

3.1.2. Congestion Avoidance………………………………………………………….……… 5

3.1.3. Timer Expiry Event……………………………………………………………….…….. 5

3.2. Tahoe……………………………………………………………………………………………. 6

3.2.1. Fast Retransmit…………………………………………………………………….…… 6

3.3. Reno……………………………………………………………………………………………… 6

3.3.1. Fast Recovery……………………………………………………………………….….. 6

3.4. New Reno……………………………………………………………………………………….. 7

3.4.1. New Reno modification to Fast Recovery………………………………………........ 7

3.5. BIC………………………………………………………………………………………………….7

3.5.1. Additive Increase……………………………………………………………………........ 8

3.5.2. Slow Start…………………………………………………………………………………. 8

3.5.3. Fast Convergence………………………………………………………………………...9

3.6. CUBIC…….………………………………………………………………………………………. 9

3.6.1. TCP-friendliness region……………..………………………………………................. 10

3.6.2. Concave region………………………………………………………………................. 11

3.6.3. Convex region………………………………………………………………………......... 11

3.6.4. Multiplicative decrease……………………………………………………………......... 11

@2017 Tetcos. This document is protected by copyright, all rights reserved 2

Tetcos Engineering

3.6.5. Fast convergence………………………...……………………………………….......... 12

4. Analysis………………………………………………………………………………………………… 13

4.1. Old Tahoe………………………………………………………………………………………… 14

4.2. Tahoe…………………………………………………………………………………………….. 15

4.3. Reno……………………………………………………………………………………………… 16

4.4. New Reno……………………………………………………………………………………….. 17

4.5. BIC………………………………………………………………………………………………… 18

4.6. CUBIC……………………………………………………………………………………………. 19

5. References……………………………………………………………………………………………. 20

@2017 Tetcos. This document is protected by copyright, all rights reserved 3

Tetcos Engineering

1. Abstract
In this paper we discuss the different congestion control algorithms of the TCP variants of Old Tahoe,
Tahoe, Reno, New Reno, BIC and CUBIC. All the algorithms suggest mechanisms for determining
when to retransmit a packet and how it should update the congestion window. In our analysis we
create a network scenario in NetSim and compare the congestion window behaviour of all the six
variants of TCP.

2. Introduction to TCP
TCP is a transport layer protocol, part of the TCP/IP suite which defines how to establish and
maintain a connection in a network. It is a connection-oriented, end-to-end reliable protocol designed
to fit into a layered hierarchy of protocols which supports a variety network applications.

TCP’s design philosophy has evolved considerably from where the goal was to develop an effective
packet switching protocol to a protocol which is fair, robust and reliable. As the internet traffic has
increased substantially over the past few decades it was important for TCP to be fair and take into
consideration the congestion in a network.

2.1 Problem of Congestion

Congestion is a situation in Communication Networks in which too many packets are present in a part
of the subnet and it subsequently leads to performance degradation. Congestion in a network may
occur when the load on the network (i.e. the number of packets sent to the network) is greater than
the capacity of the network.

Congestion Control refers to techniques that can either prevent congestion, before it happens, or
remove congestion after it has happened. TCP does this by maintaining a ‘congestion window’, which
indicates the maximum amount of data that can be sent into the network without being
acknowledged. Different variants of TCP use a certain combinations of algorithms which update the
congestion window by sensing congestion in the network and thereby preventing it.

@2017 Tetcos. This document is protected by copyright, all rights reserved 4

Tetcos Engineering

3. TCP Variants
The six variants of TCP and the congestion control algorithms used by them are mentioned below.

3.1 Old Tahoe

Old Tahoe is one of the earliest variants of TCP. It implements two algorithms called slow start and
congestion avoidance to update the congestion window.

3.1.1 Slow Start

At the start of data transmission the size of congestion window is one. This means TCP can send
only one packet until it receives an acknowledgement. When the ACK is received by the sender the
congestion window increases to two. Now the sender can send two data packets. Upon the arrival of
every new ACK the sender increases its congestion window by one. This phase is known as the slow
start phase where the congestion window increases exponentially.

So on the arrival of a new ACK, 𝑐𝑤𝑛𝑑 + = 𝑀𝑆𝑆;

3.1.2 Congestion Avoidance

TCP will continue the slow start phase until it reaches a certain threshold, or if packet loss occurs.
Now it enters in to a phase called congestion avoidance. Here the congestion window grows linearly.
This means that the congestion window increases from ‘n’ to ‘n+1’ only when it has received ‘n’ new

ACKs. The rate of growth of congestion window slows down because this is the stage where TCP is
susceptible to packet loss. The formula used here is 𝑐𝑤𝑛𝑑 += (𝑆𝑀𝑆𝑆 ∗ 𝑆𝑀𝑆𝑆)/𝑐𝑤𝑛𝑑.

3.1.3 Timer Expiry Event

TCP maintains a timer for every data packet that it has sent into the network. If it doesn’t receive an
ACK before the timer expires, it assumes that a packet loss has occurred. TCP does the following
when the timer expires

• Threshold value is reduced to half of the congestion window or twice the maximum segment
size, whichever is the maximum. 𝑠𝑠𝑡ℎ𝑟𝑒𝑠ℎ = 𝑚𝑎𝑥 (𝑐𝑤𝑛𝑑 / 2, 2 ∗ 𝑆𝑀𝑆𝑆).
• The packet which is assumed to be lost is retransmitted.

@2017 Tetcos. This document is protected by copyright, all rights reserved 5

Tetcos Engineering

• Congestion window is reduced to one.

• Slow start is resumed.

3.2 Tahoe

TCP Tahoe implements all the above mentioned algorithms used by Old Tahoe. The Fast Retransmit
algorithm was included in Tahoe to improve the response time of TCP.

3.2.1 Fast Retransmit

One of the major drawbacks of Old Tahoe is that it depends on the timer to expire before it can
retransmit a packet. TCP Tahoe tries to improve upon Old Tahoe by implementing the fast retransmit
algorithm. Fast Retransmit takes advantage of the fact that duplicate ACKs can be an indication that
a packet loss has occurred. So, whenever it receives 3 duplicate ACKs it assumes that a packet loss
has occurred and retransmits the packet.

3.3 Reno

TCP Reno retains the basic principles of Tahoe such as Slow Start, Congestion Avoidance and Fast
Retransmit. However it is not as aggressive as Tahoe in the reduction of the congestion window.
Reno implements the Fast Recovery algorithm which is described below.

3.3.1 Fast Recovery

The congestion window drop to one on the arrival of a 3 duplicate ACK can be considered as an
extreme precaution. Arrival of 3 duplicate ACKs corresponds to light congestion in the network and
there is no need for the congestion window to drop down drastically. The Fast Recovery algorithm
does the following on the arrival of a third duplicate ACK:

• The threshold value is set to half of the congestion window. 𝑠𝑠𝑡ℎ𝑟𝑒𝑠ℎ = 𝑐𝑤𝑛𝑑/2.
• The congestion window is now set to be threshold plus three times the MSS.
𝑐𝑤𝑛𝑑 = 𝑠𝑠𝑡ℎ𝑟𝑒𝑠ℎ + 3 ∗ 𝑆𝑀𝑆𝑆.
• On the arrival of another duplicate ACK the congestion window increases by one MSS. This is
done because an ACK signifies that a segment if out of the network and the sender can pump
in another packet into the network. This is somewhat similar to slow start. 𝑐𝑤𝑛𝑑 += 𝑆𝑀𝑆𝑆.

@2017 Tetcos. This document is protected by copyright, all rights reserved 6

Tetcos Engineering

• TCP remains in fast recovery phase until it receives a higher ACK from the receiver.
• On receiving a higher ACK the congestion window is set to the threshold value. From now
onwards congestion avoidance is followed. 𝑐𝑤𝑛𝑑 = 𝑠𝑠𝑡ℎ𝑟𝑒𝑠ℎ.

3.4 New Reno

Reno doesn’t improve much upon Tahoe if there are multiple packet losses in the same window. This
is because when multiple packet losses occur Reno enters fast recovery multiple times which
decreases the congestion window by half every time. In practical scenarios where multiple packet
loss in the same window is common, Reno doesn’t increase the throughput significantly. A modified
version of fast recovery algorithm was suggested to overcome this.

3.4.1 New Reno’s Modification to Fast Recovery

In the modified version of Fast Recovery, TCP stores the sequence number of the highest data
packet which is sent when the third duplicate ACK arrives in a variable called ‘recover’. New Reno
doesn’t exit fast recovery unless it has received an ACK which is higher than the recover packet.

If a new ACK arrives which has an ACK number lower than the recover packet, it immediately
retransmits that packet. This ACK is known as a partial ACK. When a partial ACK arrives the
congestion window is reduced by the amount of ‘new data’ that has been acknowledged by the partial
ACK (i.e. the amount of data that has been acknowledged after the retransmitted packet). To this one
MSS is added. In mathematical terms. 𝑐𝑤𝑛𝑑 = 𝑐𝑤𝑛𝑑 – 𝑛𝑒𝑤 𝑑𝑎𝑡𝑎 + 𝑆𝑀𝑆𝑆.

TCP stays in fast recovery mode until it receives a full ACK i.e. an ACK with a higher sequence
number than recover. Note that if there aren’t any multiple packet losses in the same window then
New Reno essentially behaves like Reno.

3.5 BIC

In BIC congestion control is viewed as a searching problem in which the system can give yes/no
feedback through packet loss as to whether the current sending rate (or window) is larger than the
network capacity. The current minimum window can be estimated as the window size at which the
flow does not see any packet loss. If the maximum window size is known, we can apply a binary
search technique to set the target window size to the midpoint of the maximum and minimum. As
increasing to the target, if it gives any packet loss, the current window can be treated as a new

@2017 Tetcos. This document is protected by copyright, all rights reserved 7

Tetcos Engineering

maximum and the reduced window size after the packet loss can be the new minimum. The midpoint
between these new values becomes a new target. Since the network incurs loss around the new
maximum but did not do so around the new minimum, the target window size must be in the middle of
the two values. After reaching the target and if it gives no packet loss, then the current window size
becomes a new minimum, and a new target is calculated. This process is repeated with the updated
minimum and maximum until the difference between the maximum and the minimum falls below a
preset threshold, called the minimum increment (Smin). This technique is called binary search
increase.

3.5.1 Additive Increase

In order to ensure faster convergence and RTT-fairness, binary search increase is combined with an
additive increase strategy. When the distance to the midpoint from the current minimum is too large,
increasing the window size directly to that midpoint might add too much stress to the network. When
the distance from the current window size to the target in binary search increase is larger than a
prescribed maximum step, called the maximum increment (Smax) instead of increasing window directly
to that midpoint in the next RTT, we increase it by Smax until the distance becomes less than Smax, at
which time window increases directly to the target. Thus, after a large window reduction, the strategy
initially increases the window linearly, and then increases logarithmically. This combination of binary
search increase and additive increase is called as binary increase. Combined with a multiplicative
decrease strategy, binary increase becomes close to pure additive increase under large windows.
This is because a larger window results in a larger reduction by multiplicative decrease and therefore,
a longer additive increase period. When the window size is small, it becomes close to pure binary
search increase – a shorter additive increase period.

3.5.2 Slow Start

After the window grows past the current maximum, the maximum is unknown. At this time, binary
search sets its maximum to be a default maximum (a large constant) and the current window size to
be the minimum. So the target midpoint can be very far. According to binary increase, if the target
midpoint is very large, it increases linearly by the maximum increment. Instead, run a “slow start”
strategy to probe for a new maximum up to Smax. So if cwnd is the current window and the maximum
increment is Smax, then it increases in each RTT round in steps cwnd+1, cwnd+2, cwnd+4,…,
cwnd+Smax. The rationale is that since it is likely to be at the saturation point and also the maximum is

@2017 Tetcos. This document is protected by copyright, all rights reserved 8

Tetcos Engineering

unknown, it probes for available bandwidth in a “slow start” until it is safe to increase the window by
Smax. After slow start, it switches to binary increase.

3.5.3 Fast Convergence

It can be shown that under a completely synchronized loss model, binary search increase combined
with multiplicative decrease converges to a fair share. Suppose there are two flows with different
window sizes, but with the same RTT. Since the larger window reduces more in multiplicative
decrease (with a fixed factor β), the time to reach the target is longer for a larger window. However,
its convergence time can be very long. In binary search increase, it takes 𝑙𝑜𝑔(𝑑) − 𝑙𝑜𝑔(𝑆𝑚𝑖𝑛) RTT
rounds to reach the maximum window after a window reduction of d. Since the window increases in a
log step, the larger window and smaller window can reach back to their respective maxima very fast
almost at the same time (although the smaller window flow gets to its maximum slightly faster). Thus,
the smaller window flow ends up taking away only a small amount of bandwidth from the larger flow
before the next window reduction. To remedy this behaviour, binary search increase is modified as
follows. After a window reduction, new maximum and minimum are set. Suppose these values are
max_wini and min_wini for flow i (i =1, 2). If the new maximum is less than the previous, this window is
in a downward trend. Then, readjust the new maximum to be the same as the new target window (i.e.
max_wini = (max_wini-min_wini)/2), and then readjust the target. After that apply the normal binary
increase. This strategy is called fast convergence.

3.6 CUBIC

In CUBIC, the window growth function is a cubic function, whose shape is very similar to the growth
function of BIC. CUBIC is designed to simplify and enhance the window control of BIC. CUBIC uses a
cubic function of the elapsed time from the last congestion event. While most alternative algorithms to
Standard TCP uses a convex increase function where after a loss event, the window increment is
always increasing, CUBIC uses both the concave and convex profiles of a cubic function for window
increase. After a window reduction following a loss event, it registers W max to be the window size
where the loss event occurred and performs a multiplicative decrease of congestion window by a
factor of β where β is a window decrease constant and the regular fast recovery and retransmit of
TCP. After it enters into congestion avoidance from fast recovery, it starts to increase the window
using the concave profile of the cubic function. The cubic function is set to have its plateau at W max so

@2017 Tetcos. This document is protected by copyright, all rights reserved 9

Tetcos Engineering

the concave growth continues until the window size becomes W max. After that, the cubic function turns
into a convex profile and the convex window growth begins. This style of window adjustment
(concave and then convex) improves protocol and network stability while maintaining high network
utilization. This is because the window size remains almost constant, forming a plateau around W max
where network utilization is deemed highest and under steady state, most window size samples of
CUBIC are close to W max, thus promoting high network utilization and protocol stability. Note that
protocols with convex growth functions tend to have the largest window increment around the
saturation point, introducing a large burst of packet losses.

The window growth function of CUBIC uses the following function:

𝑊 (𝑡) = 𝐶(𝑡 − 𝐾)3 + 𝑊𝑚𝑎𝑥 …….(1)
where C is a CUBIC parameter, t is the elapsed time from the last window reduction, and K is the
time period that the above function takes to increase W to W max when there is no further loss event
and is calculated by using the following equation:

𝐾 = 𝑟3 𝑊𝑚𝑎𝑥 𝐶 𝛽 .....…(2)

Upon receiving an ACK during congestion avoidance, CUBIC computes the window growth rate
during the next RTT period using Eq. (1). It sets W (t + RTT) as the candidate target value of
congestion window. Suppose that the current window size is cwnd. Depending on the value of cwnd.
CUBIC runs in three different modes. First, if cwnd is less than the window size that (standard) TCP
would reach at time t after the last loss event, then CUBIC is in the TCP mode. Otherwise, if cwnd is
less than W max, then CUBIC is in the concave region, and if cwnd is larger than W max, CUBIC is in the
convex region.

3.6.1 TCP-friendly region

When receiving an ACK in congestion avoidance, first check whether the protocol is in the TCP
region or not. This can be analysed by the window size of TCP in terms of the elapsed time t. Then
find the average window size of additive increase and multiplicative decrease (AIMD) with an additive
factor α and a multiplicative factor β to be the following function:

1 𝛼 2−𝛽 1
√
𝑅𝑇𝑇 2 𝛽 𝑝
……… (3)

@2017 Tetcos. This document is protected by copyright, all rights reserved 10

Tetcos Engineering

1 31
By the same analysis, the average window size of TCP with α = 1 and β = 0.5 is 𝑅𝑇𝑇 √2 𝑝 . Thus, for

Eq. 3 to be the same as that of TCP, α must be equal to 3β/2-β. If TCP increases its window by α per
RTT, then window size of TCP in terms of the elapsed time t as follows:

𝛽 𝑡
𝑊𝑡𝑐𝑝(𝑡) = 𝑊𝑚𝑎𝑥(1 − 𝛽) + 3 ………. (4)
2−𝛽 𝑅𝑇𝑇

If cwnd is less than W tcp(t), then the protocol is in the TCP mode and cwnd is set to W tcp(t) at each
reception of ACK.

3.6.2 Concave region

When receiving an ACK in congestion avoidance, if the protocol is not in the TCP mode and cwnd is
less than W max then the protocol is in the concave region. In this region, cwnd is incremented by
𝑊(𝑡+𝑅𝑇𝑇) − 𝑐𝑤𝑛𝑑
𝑐𝑤𝑛𝑑

3.6.3 Convex region

When the window size of CUBIC is larger than W max, it passes the plateau of the cubic function after
which CUBIC follows the convex profile of the cubic function. Since cwnd is larger than the previous
saturation point Wmax, this indicates that the network conditions might have been perturbed since the
last loss event, possibly implying more available bandwidth after some flow departures. Since the
Internet is highly asynchronous, fluctuations in available bandwidth always exist. The convex profile
ensures that the window increases very slowly at the beginning and gradually increases its growth
rate. This phase is called as the maximum probing phase since CUBIC is searching for a new
Wmax. Since the window increase function for the convex region is not modified, the window growth
function for both regions remains unchanged. To be exact, if the protocol is the convex region outside
𝑊(𝑡+𝑅𝑇𝑇) − 𝑐𝑤𝑛𝑑
the TCP mode, cwnd is incremented by 𝑐𝑤𝑛𝑑

3.6.4 Multiplicative decrease

When a packet loss occurs, CUBIC reduces its window size by a factor of β (set β = 0.2). A side
effect of setting β to a smaller value than 0.5 is slower convergence. The more adaptive setting of β
leads to faster convergence, it will make the analysis of the protocol much harder and also affects the
stability of the protocol. This adaptive adjustment of β is a future research issue.

@2017 Tetcos. This document is protected by copyright, all rights reserved 11

Tetcos Engineering

3.6.5 Fast Convergence

When a new flow joins the network, existing flows in the network need to give up their bandwidth
shares to allow the new flow some room for growth. To increase this release of bandwidth by existing
flows, the following mechanism called fast convergence is added. With fast convergence, when a loss
event occurs, before a window reduction of the congestion window, the protocol remembers the last
value of W max before it updates W max for the current loss event. Let us call the last value of W max to be
Wlast-max. At a loss event, if the current value of W max is less than the last value of it, W last-max, this
indicates that the saturation point experienced by this flow is getting reduced because of the change
in available bandwidth. This allows the flow to release more bandwidth by reducing W max further. This
action effectively lengthens the time for this flow to increase its window because the reduced W max
forces the flow to have the plateau earlier. This allows more time for the new flow to catch up its
window size.

@2017 Tetcos. This document is protected by copyright, all rights reserved 12

Tetcos Engineering

4. Analysis
We have used NetSim to simulate the network scenario. The scenario consists of two nodes
connected by two routers. Wired node C is the sender and wired node D is the receiver. Packets of
size 1460 bytes are generated by the application layer of the sender every 400 micro seconds of the
simulation. This amounts to 27.84 Mega Bits of data generated per second. This is higher than the
node to router link which is set to 20 Mega Bits per second. This simulates an infinite backlog of
traffic which means that TCP never runs out of data to send.

All simulations were run using NetSim version 10.0

Scenario Properties for Old Tahoe and Tahoe

Maximum Segment Size (MSS) 1460 bytes

Node – Router propagation delay 5 µs
Router – Router propagation delay 100 µs
Node – Router link speed (both) 20 Mbps
Router – Router link speed 100 Mbps
Bit Error Rate (All links) 10 e(-8)
Packet generation Rate Infinite Backlog

@2017 Tetcos. This document is protected by copyright, all rights reserved 13

Tetcos Engineering

4.1 Old Tahoe

The graph shown above is a plot of Congestion window vs. Time of Old Tahoe for the scenario
shown above. Each point on the graph represents the congestion window at the time when the
packet is sent. You can observe that after every congestion window drop (caused by packet loss or
timer expiry) Old Tahoe enters slow start, thereby increasing its window exponentially and then
enters congestion avoidance where it increases the congestion window linearly.

@2017 Tetcos. This document is protected by copyright, all rights reserved 14

Tetcos Engineering

4.2 Tahoe

Tahoe uses the fast retransmit algorithm with which it responds to packet errors faster than Old
Tahoe. Comparing the graphs of Old Tahoe and Tahoe one can observe that the latter drops the
congestion window and retransmits faster than Old Tahoe (when three duplicate ack’s are received).

Scenario Properties for Reno and New Reno

Maximum Segment Size (MSS) 1460 bytes

Node – Router propagation delay 5 µs
Router – Router propagation delay 100 µs
Node – Router link speed (both) 20 Mbps
Router – Router link speed 100 Mbps
Bit Error Rate (All links) 10 e(-7)
Packet generation Rate Infinite Backlog

@2017 Tetcos. This document is protected by copyright, all rights reserved 15

Tetcos Engineering

4.3 Reno

TCP Reno upon receiving the third duplicate ACK sets its congestion window according to the
formula cwnd = cwnd/2 + 3*MSS. On further arrival of duplicate ACKs it increases its congestion
window by one MSS. If it receives a higher ACK then it drops the congestion window to the new
threshold value. This mechanism is known as fast recovery.

@2017 Tetcos. This document is protected by copyright, all rights reserved 16

Tetcos Engineering

4.4 New Reno

In the New Reno’s modification to fast recovery, TCP doesn’t exit the fast recovery phase until it
receives a full ACK (the ACK for the segment which was sent just before the arrival of the third
duplicate ACK). If it receives a duplicate ACK in the fast recovery phase it retransmits the segment
corresponding to the duplicate ACK. It then increases the congestion window by one MSS. Hence we
see a sharper rise in the congestion window size as New Reno doesn’t exit the fast recovery phase
until it receives the full ACK.

Scenario Properties for BIC and CUBIC

Maximum Segment Size (MSS) 1460 bytes

Node – Router propagation delay 5 µs
Router – Router propagation delay 10000 µs
Node – Router link speed (both) 20 Mbps
Router – Router link speed 100 Mbps
Bit Error Rate (All links) 10 e(-8)
Packet generation Rate Infinite Backlog

@2017 Tetcos. This document is protected by copyright, all rights reserved 17

Tetcos Engineering

4.5 BIC

The graph shown above is a plot of Congestion window vs. Time of BIC for the scenario shown
above. Each point on the graph represents the congestion window at the time when the packet is
sent. You can observe Binary Search, Additive Increase, Fast Convergence, Slow Start phases in the
above graph.

@2017 Tetcos. This document is protected by copyright, all rights reserved 18

Tetcos Engineering

4.6 CUBIC

The graph shown above is a plot of Congestion window vs. Time of CUBIC for the scenario shown
above. You can observe Multiplicative decrease, Fast Convergence, Concave and Convex regions in
the above graph.

@2017 Tetcos. This document is protected by copyright, all rights reserved 19

Tetcos Engineering

5. References
1. Information Sciences Institute, “Transmission Control Protocol” RFC 793, September 1981
2. W. Stevens, “TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery
Algorithms,” RFC2001, January 1997.
3. M. Allman, V. Paxson, W. Stevens, “TCP Congestion Control”, RFC 2581, April 1999.
4. S. Floyd, T. Henderson “The New Reno modification to TCP’s Fast Recovery Algorithm”, RFC
2582, April 1999.
5. Lisong Xu, Khaled Harfoush, and Injong Rhee, “Binary Increase Congestion Control for Fast,
Long Distance Networks” IEEE INFOCOM. 2004，Vol 4: pp.2514-2524
6. Sangtae Ha, Injong Rhee, Lisong Xu , “CUBIC: A New TCP-Friendly High-Speed TCP Variant”

@2017 Tetcos. This document is protected by copyright, all rights reserved 20