Rucete ✏ AP Computer Science Principles In a Nutshell
5. Computer Systems and Networks
This chapter introduces computing devices and systems, the structure of the internet, packet transmission, fault tolerance, Internet protocols (IP, TCP, UDP), redundancy, efficiency of sequential vs. parallel computing, and the role of distributed computing in solving complex problems.
Computing Devices and Systems
• A computing device is any physical object capable of running programs (e.g., computers, smartphones, routers, smart sensors).
• Input devices include keyboards, mice, scanners; output devices include monitors, printers, speakers.
• A computing system is a group of devices working together, often forming a network.
Computer Networks
• A network connects devices to send and receive data, using bandwidth (data transmitted per unit time) as a key measure.
• A path is a sequence of devices connecting sender to receiver; routing selects a path.
• If a path fails, routing reroutes data to maintain communication (fault tolerance).
The Internet
• The internet is a physical network made of fiber optics, cables, and wireless technologies.
• Devices communicate using open standard protocols for interoperability and scalability.
• Scalability enables the internet to grow and meet new demands without collapsing.
Internet Protocol (IP)
• IP handles addressing and routing of online communications.
• IPv4 uses 32-bit addresses (about 4.3 billion addresses); IPv6 uses 128-bit addresses (almost infinite unique addresses).
• Devices can dynamically change IP addresses when reconnecting to the internet.
Transmission Control Protocol (TCP)
• TCP ensures reliable data transfer by splitting data into packets and confirming their arrival.
• Each packet contains sender and receiver IP addresses, a packet number, and data.
• If packets are lost, TCP requests retransmission.
User Datagram Protocol (UDP)
• UDP allows faster data transmission without confirming packet arrival.
• Useful for live video streaming and gaming where speed is prioritized over reliability.
Fault Tolerance and Redundancy
• Redundancy in a network means having multiple pathways for data transmission.
• Fault tolerance: The ability of a system to continue operating properly even if some components fail.
• The internet’s redundant design allows rerouting data through alternate paths if parts of the network fail.
• More redundancy increases reliability but may also increase costs and complexity.
Internet vs. World Wide Web
• Internet: The global network of interconnected computing devices (physical layer).
• World Wide Web (WWW): A system of linked pages, files, and resources accessed via the internet (application layer).
• Web browsers retrieve and display pages using protocols like HTTP and HTTPS.
Sequential vs. Parallel Computing
• Sequential computing: Completing tasks one after another.
• Parallel computing: Breaking a problem into sub-tasks solved simultaneously across multiple processors.
• Parallel computing increases speed and efficiency but requires coordination and may introduce overhead.
• Speedup = time for sequential solution ÷ time for parallel solution.
Distributed Computing
• Distributed computing: Using multiple, often geographically dispersed, devices to solve a single problem collaboratively.
• Examples include SETI@home (searching for extraterrestrial signals) and Folding@home (protein folding research).
• Distributed systems improve processing power, reliability, and storage capacity beyond what a single device could achieve.
Benefits and Challenges of Distributed Systems
• Benefits:
• Scalability: Systems grow easily by adding more devices.
• Fault Tolerance: Failure of individual nodes does not cripple the system.
• Resource Sharing: Devices share data, computation, and storage resources.
• Challenges:
• Complexity: Managing distributed systems is harder than single-device systems.
• Synchronization: Ensuring all parts work together smoothly can be difficult.
In a Nutshell
Modern computing relies on networks of devices and layered protocols like IP, TCP, and UDP to transfer and manage data efficiently and reliably. Redundancy and fault tolerance ensure robust network operations. Parallel and distributed computing methods allow large, complex problems to be solved more quickly and effectively by harnessing the combined power of many processors and systems.