In this paper we describe our experience in the design and teaching of a new evolutionary introductory programming course in a new Distributed Computing and Communications Systems Technology program at the University of Guelph-Humber. This course is evolutionary and innovative because it integrates the use of HTML, JavaScript, and Java in a one-semester introductory computer programming course. This is a marked departure from the use of a single conventional, general purpose, programming language such as Java or C++. The course is designed with two goals in mind: to improve the students experience in their first computer programming; and to achieve retention in the new program.

We present a deflection method suitable for Manhattan-street networks (MSN), which, without dropping packets, limits the number of hops travelled by a packet on its way from source to destination. The proposed routing scheme is intended for bidirectional networks-with four incoming and four outgoing links per switch. In terms of average performance measures, our method performs no worse than the best local routing schemes, when the offered load is light or moderate. Unlike other deflection schemes proposed for MSN, our method is inherently asynchronous. This simplifies the switch design and eliminates some problems with the original synchronous approach, like the need to account for slight discrepancies in transmission rates of different switches.

This paper proposes an efficient, fair, and simple capacity-1 protocol replacing FDDI. In the recent years, two network protocols gained predominance in very fast local and in metropolitan area networks: FDDI and DQDB. The FDDI ring-protocol combination is well-suited for large local ("campus") area networks. However, for larger networks, FDDI users suffer excessive access delays, even under very light traffic. Moreover, FDDI achieves full capacity only in special circumstances, and heavy load is not a sufficient condition for achieving it. We propose a network topology that is very similar to the dual ring and a token-passing protocol that yield a true capacity-1 network even for modest values of the TTRT. Its additional properties include total fairness and very small access delays under light load. This makes the proposed protocol a reasonable option to use in combination with the dual counter-rotating ring. The proposed network should not be difficult to implement using standard FDDI components.>

The authors present a simple MAC-level protocol for high-speed ring networks. The protocol is fair and its maximum achievable throughput does not deteriorate with the increasing propagation length of the ring. The proposed protocol operates on the so-called 'spiral ring' topology, including the single-segment regular ring as a special case. It accommodates synchronous traffic automatically, without any bandwidth preallocation or similar special efforts, yet with an arbitrarily low jitter and without packet loss.>

A token ring protocol which restricts the transmission rights to the station possessing the token is not a good candidate for driving a gigabit network. As the network's transmission rate becomes higher, the impact of the token transition time (which is idle) becomes more pronounced and the effective throughput of the network becomes a smaller and smaller fraction of the nominal channel capacity. The authors present a family of token passing protocols for ring networks which are devoid of this unpleasant property. The protocols are simple and inexpensive, yet they possess a number of advantageous properties: high capacity (similar to METARING and independent of the propagation length of the ring), fairness, and natural accommodation of synchronous and isochronous traffic.>

Contention control strategies for high speed backbone networks are studied. Packet loss is a serious problem in a backbone scenario due to the speed mismatch of the different networks. The authors' new mechanism tries to reduce the packet loss ratio for fiber distributed data interface (FDDI) operating as a backbone. When one buffer is overloaded and packets would be lost, the new schemes try to store those packets at another station. The proposed mechanism provides significant reduction in packet loss and prevents congestion in other parts of an interconnected network.

A new MAN/WAN medium-access control (MAC) protocol called CBRMA++/SR is introduced. It is suitable for a folded bus topology with optical fiber used as the transmission medium. The eminent features of CBRMA++/SR are total fairness and full bandwidth utilization. The preemptive approach employed in bandwidth allocation enables CBRMA+++/SR not to suffer from the deficiencies of DQDB, which stem mostly from the inconsistency of the distributed queue at a given instant. The protocol is based on the concept of fair share, which also provides a solid ground to support connection-oriented services. A powerful slot-reuse scheme complements the protocol.>

The authors propose an efficient, fair, and simple capacity-1 protocol that outperforms both fiber distributed data interface (FDDI) and distributed queue dual bus (DQDB). In the recent years, two network protocols gained predominance in very fast local and in metropolitan area networks: FDDI (using a single or dual ring) and DQDB (using a dual bus). The authors propose a network topology that is very similar to the dual ring of FDDI and a token-passing protocol suitable for this topology. Their solution yields a true capacity-1 network with zero access delay under light load. Moreover, it naturally accommodates synchronous traffic and is absolutely fair. The proposed protocol has a much better performance than FDDI; it also outperforms DQDB for sufficiently fast (or long) networks.>

A possible network architecture suitable for a next-generation follow-up to FDDI Is presented. The solution can be scaled-up to arbitrary transmission rates (limited only by the technological barriers) and arbitrarily long networks. The protocol allows every station to transmit when the ring is silent, even when the station is not in possession of the token. The resulting token-passing protocol yields a capacity-1 network irrespective of the value of TTRT. Additionally, stations endure a very small access delay under light load. The need to single out synchronous traffic is eliminated, guaranteeing each station the right to transmit at fixed intervals. This modification allows graceful station insertion and deletion; moreover, the recovery of a lost token can be done on the fly without disrupting the operation of the network.>

The paper demonstrates that the performance of MAC-level protocols is not very meaningful, when considered alone. The true object of performance analysis should be the pair (T,P), where T stands for network topology, and P for MAC-level protocol. The paper examines the performance of two MAC-level protocols: DQDB and LCSMA-CD/U/P. As both of them are capacity-1 protocols, they achieve the same maximum throughput; thus, the focus is on the packet access time and on fairness. The investigation shows that either protocol can be shown as greatly unfair or totally fair, depending on the topology that it operates on. As a byproduct, it is shown that the UU-BUS topology is quite interesting and merits further studies.>

A collision protocol called piggyback Ethernet which is applicable for fast bus-type networks based on single, uniform, unsegmented, broadcast-type carriers is presented. Although piggyback Ethernet assumes that each station possesses certain global knowledge about the geometry of the network, it is argued that this knowledge can be acquired automatically at a reasonably small expense. The performance of the protocol for very heavy traffic is not worse than the performance of other protocols operating in a completely synchronized manner. The advantage this solution is its very good performance for light loads which is, for practical purposes, not worse than the performance of Ethernet. Moreover, compared to other protocols for fast networks, piggyback Ethernet has quite modest hardware requirements. The network is based on a straightforward bidirectional cable, each station has only one connection to the bus, and the actual length of the carrier is not bigger than the maximum distance between two stations.>

External merge sorting consists of two phases; the formation of initial runs and the merge of the runs into a sorted file (possible in several merge passes). The replacement selection algorithm is typically used to create initial runs. It employs two levels of memory: main memory and a sequential accessed tape-like memory. Natural selection uses a third memory level to increase the average length of an initial run. Although natural selection produces longer runs than replacement selection, it is not competitive because the added overhead is excessive. This paper discusses two other methods of creating longer initial runs in a 3-level memory. Their efficiency is higher than that of natural selection and they also outperform a 3-level replacement selection in most circumstances. The second algorithm has the additional feature that it allows the auxiliary memory to be broken into several smaller parts, thus increasing the average run length. As most computer systems have rotating mass storage, this property is of practical value.