Upd | Y2down App

5-page conference paper — "Y2Down: A Privacy-Preserving, Lightweight YouTube Downloader for Mobile" (outline + abstract + section summaries) Title: Y2Down: A Privacy-Preserving, Lightweight YouTube Downloader for Mobile Abstract: Y2Down is a lightweight mobile application designed to let users download video and audio streams from public video platforms while minimizing privacy risks and complying with platform terms where possible. This paper presents the app's architecture, offline-first design, adaptive download strategies, codec-aware transcoding, and privacy-preserving telemetry. We evaluate Y2Down's performance on mid-range Android devices, measure bandwidth and battery impact, and discuss legal and ethical considerations. Results show Y2Down can download and transcode 10-minute 1080p videos with median CPU usage of 18% and battery drain of 6% for a single download, while preserving user anonymity through on-device processing and minimized network metadata exposure.

Introduction

Motivation: offline viewing demand, unreliable connectivity, data caps. Challenges: platform restrictions, varying stream formats, device constraints, privacy concerns. Contributions: (1) modular downloader with adaptive bitrate selection, (2) on-device transcoding pipeline optimized for mobile, (3) privacy-first network design, (4) evaluation on real devices.

Background and Related Work

Video streaming formats (DASH, HLS), muxing/container issues. Existing downloaders and libraries (yt-dlp, youtube-dl, mpv), mobile apps. Mobile transcoding frameworks (FFmpeg, MediaCodec). Privacy research on metadata leakage and telemetry.

Design Goals

Privacy: minimize external leaks, anonymize nonessential telemetry. Efficiency: low CPU, memory, battery footprint. Compatibility: support multiple container/codec combos and adaptive streams. Usability: simple UI, queueing, resume support. y2down app upd

Architecture

High-level components: Fetcher, Manifest Parser, Segment Downloader, Adaptive Selector, Transcoder, Storage Manager, UI/Queue, Telemetry Guard. Offline-first: manifest and segment caching; resume via persistent state. Network strategy: fetch manifests via short-lived proxies (optional), DNS-based minimal requests, segmented parallelism with concurrency limits.

Implementation

Platform: Kotlin + Jetpack Compose for Android; light iOS sketch. Fetcher: HTTP client with connection pooling and per-host rate-limiting. Manifest Parser: supports DASH/HLS, extracts representations, segment maps. Adaptive Selector: chooses format by device capabilities, battery state, and user preference (quality vs. size). Transcoding: MediaCodec hardware-accelerated pipeline with fallback to software FFmpeg; container assembly with mp4/mkv muxer. Storage: file-per-download with temporary .part state, resume tokens. Privacy measures: on-device parsing/transcoding, minimal external calls, telemetry opt-in, local-only logs.

Evaluation