English
Let us consider that host `A` uses a window of three segments. It thus sends three back-to-back segments at `10 Mbps` and then waits for an acknowledgment. Host `A` stops sending segments when its window is full. These segments reach the buffers of router `R1`. The first segment stored in this buffer is sent by router `R1` at a rate of `2 Mbps` to the destination host. Upon reception of this segment, the destination sends an acknowledgment. This acknowledgment allows host `A` to transmit a new segment. This segment is stored in the buffers of router `R1` while it is transmitting the second segment that was sent by host `A`... Thus, after the transmission of the first window of segments, the reliable transport protocol sends one data segment after the reception of each acknowledgment returned by the destination. In practice, the acknowledgments sent by the destination serve as a kind of `clock` that allows the sending host to adapt its transmission rate to the rate at which segments are received by the destination. This `self-clocking` is the first mechanism that allows a window-based reliable transport protocol to adapt to heterogeneous networks [Jacobson1988]_. It depends on the availability of buffers to store the segments that have been sent by the sender but have not yet been transmitted to the destination.
However, transport protocols are not only used in this environment. In the global Internet, a large number of hosts send segments to a large number of receivers. For example, let us consider the network depicted below which is similar to the one discussed in [Jacobson1988]_ and :rfc:`896`. In this network, we assume that the buffers of the router are infinite to ensure that no packet is lost.
If many senders are attached to the left part of the network above, they all send a window full of segments. These segments are stored in the buffers of the router before being transmitted towards their destination. If there are many senders on the left part of the network, the occupancy of the buffers quickly grows. A consequence of the buffer occupancy is that the round-trip-time, measured by the transport protocol, between the sender and the receiver increases. Consider a network where 10,000 bits segments are sent. When the buffer is empty, such a segment requires 1 millisecond to be transmitted on the `10 Mbps` link and 5 milliseconds to be the transmitted on the `2 Mbps` link. Thus, the measured round-trip-time measured is roughly 6 milliseconds if we ignore the propagation delay on the links. If the buffer contains 100 segments, the round-trip-time becomes :math:`1+100 \times 5+ 5` milliseconds as new segments are only transmitted on the `2 Mbps` link once all previous segments have been transmitted. Unfortunately, if the reliable transport protocol uses a retransmission timer and performs `go-back-n` to recover from transmission errors it will retransmit a full window of segments. This increases the occupancy of the buffer and the delay through the buffer... Furthermore, the buffer may store and send on the low bandwidth links several retransmissions of the same segment. This problem is called `congestion collapse`. It occurred several times during the late 1980s on the Internet [Jacobson1988]_.
The `congestion collapse` is a problem that all heterogeneous networks face. Different mechanisms have been proposed in the scientific literature to avoid or control network congestion. Some of them have been implemented and deployed in real networks. To understand this problem in more detail, let us first consider a simple network with two hosts attached to a high bandwidth link that are sending segments to destination `C` attached to a low bandwidth link as depicted below.
To avoid `congestion collapse`, the hosts must regulate their transmission rate [#fcredit]_ by using a `congestion control` mechanism. Such a mechanism can be implemented in the transport layer or in the network layer. In TCP/IP networks, it is implemented in the transport layer, but other technologies such as `Asynchronous Transfer Mode (ATM)` or `Frame Relay` include congestion control mechanisms in lower layers.
Let us first consider the simple problem of a set of :math:`i` hosts that share a single bottleneck link as shown in the example above. In this network, the congestion control scheme must achieve the following objectives [CJ1989]_ :
The congestion control scheme must `avoid congestion`. In practice, this means that the bottleneck link cannot be overloaded. If :math:`r_i(t)` is the transmission rate allocated to host :math:`i` at time :math:`t` and :math:`R` the bandwidth of the bottleneck link, then the congestion control scheme should ensure that, on average, :math:`\forall{t} \sum{r_i(t)} \le R`.
The congestion control scheme must be `efficient`. The bottleneck link is usually both a shared and an expensive resource. Usually, bottleneck links are wide area links that are much more expensive to upgrade than the local area networks. The congestion control scheme should ensure that such links are efficiently used. Mathematically, the control scheme should ensure that :math:`\forall{t} \sum{r_i(t)} \approx R`.
The congestion control scheme should be `fair`. Most congestion schemes aim at achieving `max-min fairness`. An allocation of transmission rates to sources is said to be `max-min fair` if :
no link in the network is congested
the rate allocated to source :math:`j` cannot be increased without decreasing the rate allocated to a source :math:`i` whose allocation is smaller than the rate allocated to source :math:`j` [Leboudec2008]_ .
Depending on the network, a `max-min fair allocation` may not always exist. In practice, `max-min fairness` is an ideal objective that cannot necessarily be achieved. When there is a single bottleneck link as in the example above, `max-min fairness` implies that each source should be allocated the same transmission rate.
To visualize the different rate allocations, it is useful to consider the graph shown below. In this graph, we plot on the `x-axis` (resp. `y-axis`) the rate allocated to host `B` (resp. `A`). A point in the graph :math:`(r_B,r_A)` corresponds to a possible allocation of the transmission rates. Since there is a `2 Mbps` bottleneck link in this network, the graph can be divided into two regions. The lower left part of the graph contains all allocations :math:`(r_B,r_A)` such that the bottleneck link is not congested (:math:`r_A+r_B<2`). The right border of this region is the `efficiency line`, i.e. the set of allocations that completely utilize the bottleneck link (:math:`r_A+r_B=2`). Finally, the `fairness line` is the set of fair allocations.
Possible allocated transmission rates
As shown in the graph above, a rate allocation may be fair but not efficient (e.g. :math:`r_A=0.7,r_B=0.7`), fair and efficient ( e.g. :math:`r_A=1,r_B=1`) or efficient but not fair (e.g. :math:`r_A=1.5,r_B=0.5`). Ideally, the allocation should be both fair and efficient. Unfortunately, maintaining such an allocation with fluctuations in the number of flows that use the network is a challenging problem. Furthermore, there might be several thousands flows that pass through the same link [#fflowslink]_.
To deal with these fluctuations in demand, which result in fluctuations in the available bandwidth, computer networks use a congestion control scheme. This congestion control scheme should achieve the three objectives listed above. Some congestion control schemes rely on a close cooperation between the end hosts and the routers, while others are mainly implemented on the end hosts with limited support from the routers.
A congestion control scheme can be modeled as an algorithm that adapts the transmission rate (:math:`r_i(t)`) of host :math:`i` based on the feedback received from the network. Different types of feedback are possible. The simplest scheme is a binary feedback [CJ1989]_ [Jacobson1988]_ where the hosts simply learn whether the network is congested or not. Some congestion control schemes allow the network to regularly send an allocated transmission rate in Mbps to each host [BF1995]_.
Let us focus on the binary feedback scheme which is the most widely used today. Intuitively, the congestion control scheme should decrease the transmission rate of a host when congestion has been detected in the network, in order to avoid congestion collapse. Furthermore, the hosts should increase their transmission rate when the network is not congested. Otherwise, the hosts would not be able to efficiently utilize the network. The rate allocated to each host fluctuates with time, depending on the feedback received from the network. The figure below illustrates the evolution of the transmission rates allocated to two hosts in our simple network. Initially, two hosts have a low allocation, but this is not efficient. The allocations increase until the network becomes congested. At this point, the hosts decrease their transmission rate to avoid congestion collapse. If the congestion control scheme works well, after some time the allocations should become both fair and efficient.
Evolution of the transmission rates
Various types of rate adaption algorithms are possible. `Dah Ming Chiu`_ and `Raj Jain`_ have analyzed, in [CJ1989]_, different types of algorithms that can be used by a source to adapt its transmission rate to the feedback received from the network. Intuitively, such a rate adaptation algorithm increases the transmission rate when the network is not congested (ensure that the network is efficiently used) and decrease the transmission rate when the network is congested (to avoid congestion collapse).